fix(workspace): remove the 1-5 prompt rating from the workspace (#447)

The star is only for favorites (maintainer ruling 2026-10-28). The workspace
no longer shows or edits a prompt rating: no Rating field in the Library
editor, no stars on Library cards, no "Highest rated" sort on Images or the
Library, and no rating column or rated-images total in System > Statistics.

The editor never sends `rating`, so a workspace save keeps a stored rating
as it is. A saved view or Library visit with the retired rating sort opens on
Recently added. Data and API are unchanged; legacy pages keep their ratings.

Recorded as deviation D153 (amends D14, D23, D96); concept comparisons map
the concept through it.

TESTING.md

@@ -731,8 +731,9 @@ The shared inline and collection inspector reads the current rooted file's suppo
 embedded records, exact generation facts and recorded or inferred prompt evidence.
 It renders in two concept surfaces (#249 D20–D22, D38). The Images side panel shows
 only the concept content: cover, title, filename, the evidence badge with a category
-chip naming its origin (D146), the + and − prompt blocks with copy and send, tag chips (the file's own image
-tags first, D88, then the linked tags; image tags and the linked positive's own tags
+chip naming its origin (D146), the + and − prompt blocks with copy and send, the tag chips in labelled
+sections after the negative prompt (D154: "Image tags · N" for the file's own image tags, D88, then "Prompt
+tags · N" or "Recipe tags · N" for the linked tags; image tags and the linked positive's own tags
 filter Images, D21), the six facts, and a footer with "Send to ComfyUI" and
 "Prompt history" only when prompt roles resolved or a link is saved; the sticky panel
 scrolls inside the window. The viewer's details (`#viewer-details`) repeat that
@@ -1087,7 +1088,8 @@ Recorded shell deviations:
   palette, then the OS.
 - **D5**: retired by Activity (#249 Activity-1). The concept's Activity button and
   Activity row are shown; their differences are D52 to D54.
-- **D28**: no COLLECTIONS, and no Design review. The bottom nav is
+- **D28**: no Design review. (Amended by #331: the COLLECTIONS group ships; its rows,
+  counts and phone layout are D136.) The bottom nav is
   the concept's Activity and System, plus the production Overview hub (its own
   outline home glyph, not the concept's `activity` pulse, and no count).
   The connection reads "Your library, on your server" with a neutral muted dot.
@@ -1384,8 +1386,8 @@ The Images page renders the concept's DOM with real data. Its controls, top to b
 - **Toolbar** (`toolbar.mjs`, shared with the Library in PR E): `#search` "Search
   prompts, tags, filenames…" (debounced 250 ms; filename, linked positive text and
   Library title, and tag names), `#category-filter` "All categories" plus the
-  Library's categories, `#sort` (Recently added, Name A–Z, Highest rated, then Oldest
-  first and Name Z–A), and `Filters` (opens "Refine this view").
+  Library's categories, `#sort` (Recently added, Name A–Z, then Oldest first and Name
+  Z–A; no Highest rated, D153), and `Filters` (opens "Refine this view").
 - **Filter row**: Grid / List / Masonry (`.segmented`, group "Image layout"),
   "Thumbnail size" Comfortable / Large / Compact, the scope chip ("All output folders"
   or "<source> · <folder>"), and "Group by positive prompt →" (goes to the Library).
@@ -1395,7 +1397,7 @@ The Images page renders the concept's DOM with real data. Its controls, top to b
   (click, Enter, Space); Info selects the card into the side panel; the star toggles
   the favorite; arrows, Home and End move between cards. Bounded rendering keeps the
   DOM small, and returning to Images restores filters, layout, size, selection and
-  scroll.
+  scroll (a layout or size the toolbar never changed follows newer defaults, D157).
 - **Continuation** (`#gallery-continuation`): scrolling near it loads the next page;
   `Load more images` and `Retry` do the same by hand, with a toast per batch.
 - **Viewer** (`dialog#viewer-dialog`): "<title> · <filename>", the concept toolbar
@@ -1411,7 +1413,7 @@ How the earlier workspace Images filters moved (old → new). The legacy `/galle
 | Source | "Refine this view" → Source (D14); shown in the scope chip |
 | Folder | "Refine this view" → Folder (D14); shown in the scope chip |
 | Media (All media / Images / Videos) | "Refine this view" → Media (D14) |
-| Sort: Newest first / Oldest first / Filename A–Z / Filename Z–A | toolbar sort: Recently added / Oldest first / Name A–Z / Name Z–A, plus Highest rated (D13, D14) |
+| Sort: Newest first / Oldest first / Filename A–Z / Filename Z–A | toolbar sort: Recently added / Oldest first / Name A–Z / Name Z–A (D13, D14; no Highest rated, D153) |
 | (none) | toolbar category (the effective category, D141) and "Refine this view" → Tags, through the linked positive (D13, D14) |
 | Refresh | Import → "Scan an output folder" → Rescan (D15) |
 | Grid / List / Masonry, Comfortable / Large / Compact | the same choices in the concept filter row |
@@ -1429,8 +1431,10 @@ Recorded deviations on this page (spec §8):
   "Removed from favorites." and a rollback toast on failure.
 - **D8**: every Images card has the concept's select checkbox before the star; the
   `#bulkbar` shows "N selected" and only the real actions in the concept's order (Add
-  tags, then D85's Auto-tag selected, Set category, Export metadata, Clear). Add to
-  collection waits for collections; "Remove links…" is dropped. (Amended by #249 PR D:
+  tags, then D85's Auto-tag selected, Set category, Add to collection, Export metadata,
+  Clear); "Remove links…" is dropped. (Amended by #331: Add to collection ships, and
+  Remove from collection follows it while a collection filter is active, see D138.)
+  (Amended by #249 PR D:
   Add tags ships, see "Bulk tagging from the selection bar"; #392 adds range, keyboard,
   touch and "all N matching" selection, see "Selecting many images", D135.) The selection survives Load more and
   bounded re-mounting and is cleared whenever the collection is replaced (search,
@@ -1446,9 +1450,9 @@ Recorded deviations on this page (spec §8):
   checked images' own categories: it opens the concept's `simpleInput` dialog
   ("Categorize selected images", `input#action-input` labelled Category) and writes
   `PUT`/`DELETE /prompt_manager/images/category`, never a prompt (see D140). Prompt
-  categories are changed in the Prompt library editor and on the legacy admin page's
-  "Set Category"; `POST /prompt_manager/bulk/category` is unchanged and serves that
-  legacy page only. The write carries no page signal: leaving Images while it runs lets
+  categories are changed in the Prompt library editor, the Prompt library's selection
+  bar (D148, #437) and on the legacy admin page's "Set Category"; both bulk ones write
+  `POST /prompt_manager/bulk/category`. The write carries no page signal: leaving Images while it runs lets
   it finish without a toast, and the return loads the collection afresh instead of the
   saved visit. Its success reloads the categories and the listing together (clearing
   the selection, focus to the heading, and, as any replaced collection does, the
@@ -1520,6 +1524,9 @@ Recorded deviations on this page (spec §8):
   filters, while recipe, recorded-run and embedded tags stay plain `span.chip`.
   Amended by D88 (#249 PR C): the file's own current image tags lead the same row
   and are filters too; a linked tag already shown (case-insensitive) is left out.
+  Amended by D154 (#449): the image tags and the linked tags each sit in their own
+  labelled section, and a linked tag stays in its section even when an image tag
+  has the same name.
   Amended by D146 (#332): the category chip also names its origin in visible text
   after the category (" · this image", " · prompt", " · recipe"), and an image's own
   category shows for an unlinked or ambiguous file too.
@@ -1679,6 +1686,29 @@ Recorded deviations on this page (spec §8):
   `scroll-margin-block-end: 96px`, so it is never under the toast.
   `Rescan toast does not cover the discovery alert it leaves` ends the window inside
   the toast's band at 1600 and 390 pixels and measures both boxes.
+- **D157** (#451, maintainer request, beyond the concept): Images opens in **Masonry**
+  when this browser has no Default layout (`gallery-settings.js`
+  `DEFAULT_SETTINGS.workspaceLayout`); the concept starts at Grid (`state.view`, app.js
+  208). A stored choice is kept, and an invalid value reads as Masonry. Saved views,
+  favorites, collections and the embedded galleries keep their fixed Grid/Comfortable.
+  The return to Images (the last visit) keeps a layout or thumbnail size its toolbar
+  changed; a value it kept from the defaults it started with follows a Default layout or
+  Thumbnail size saved since, in this tab or another (a toolbar List then Default =
+  Masonry stays List; a toolbar List then back to Grid counts as untouched). A cover crop
+  of a generated image (Grid and List card art, Library card art, the side panel's and
+  History's inspector cover, the Tag image preview) uses `object-position: 50% 20%`
+  instead of the concept's centre: it acts only when the image is taller than its box
+  (portraits keep heads and faces), and a wider image stays centred. Masonry shows the
+  whole image and keeps the centre. The concept comparisons (`openProduction`) and the
+  specs of the Grid's own geometry start production from the concept's state, a stored
+  Default layout of Grid (`startInGrid`, concept-inventory.mjs, only while the browser
+  stored nothing); the comparisons do not measure `object-position`.
+  `a fresh browser opens Images in Masonry (#451)`, `a Default layout saved later reaches
+  the next Images visit in this tab (#451)`, `... in another tab ... (#451)` and `cover
+  crops keep the top of a portrait image; Masonry stays centred (#451)`
+  (workspace-system.spec.mjs) cover it. A browser whose legacy gallery wrote the old
+  default (`"grid"`) keeps Grid until Default layout is changed or "Reset display
+  defaults" is saved.
 
 `tests/browser/inspector-concept.spec.mjs` compares the side panel and the viewer's
 full details (#249 D4) for alpine_00001.png, at 1600x1100 and 390x844 in both themes:
@@ -1784,6 +1814,30 @@ Inspect plan keeps D60–D66 (plan D60 → D83, D61 → D84, D65 → D88, D68
   concept image's tags saved as its image tags production still equals the concept.
   After a Tag save the selected file's inspection is re-read, so the panel redraws
   the row without reselecting and the viewer, which shares that inspection, shows it.
+  Amended by D154 (#449): the one row is now two labelled sections.
+- **D154** (#449, beyond the concept): the inspector's tags no longer follow the
+  negative prompt as a bare `.tags` row, where they read as part of it. Every surface of
+  the shared inspector (the Images side panel, the viewer details, History's inline
+  current file, the Library-embedded gallery and Activity) draws them in
+  `section.tag-section` regions in the concept's prompt-section frame (the
+  `.prompt-section` top rule with 15px padding and margin, a `.prompt-heading` with a
+  `.role-label`, the row's own 12px top margin dropped; workspace.css). The label is
+  muted (`--ca-muted`, as `.meta-grid dt`), because tags are neither prompt role. First
+  `section.tag-section.image-tags` "Image tags · N" (the file's current image tags, D88),
+  then `section.tag-section.linked-tags` "Prompt tags · N" for the linked positive's
+  tags or "Recipe tags · N" for a recipe's own (`tags_scope`, named as D146's category
+  chip names its origin). Each region is named without its count and with the prompt
+  blocks' context ("Current file image tags" in History). An empty source draws no
+  section. Repeats are removed within a source only, so a linked tag that is also an
+  image tag shows in both sections and each count is true. D21's filter rules are
+  unchanged. History's current-file read carries no `image_tags` facet (D88), so there
+  only the linked section shows. `tests/browser/inspector-concept.spec.mjs` wraps the
+  concept's row in the same sections at runtime (`withTagSections`, the D154 rule
+  injected) and compares the measured sections (frame, heading, label, row); a
+  mutation of the label color, the row margin or the top rule fails it. Browser tests:
+  "D154: image tags sit in their own labelled section after the negative prompt, in
+  the panel and the viewer" and "the inline current file labels its image tags after
+  its negative prompt".
 - **D115** (beyond the plan's list): the concept always has a selected image, so
   its "Auto-tag images" always opens the dialog. Production opens it for the image
   selected in Images when it was last left; with none it is `aria-disabled` and
@@ -1869,6 +1923,7 @@ production-versus-concept difference of PR C and its record:
 | Suggestions count field | D91 (ruling T3-bulk) |
 | No prototype note, "Suggestions from …", no " in this tab" | D29 |
 | Image tags first in the inspector and viewer chip row | D88 (ruling T7/C2) |
+| Image tags and linked tags in labelled sections after the negative prompt | D154 (#449) |
 | Auto-tag images `aria-disabled` without a selected image | D115 |
 | Retry leads the actions when the controller allows it | D116 (ruling T5) |
 | States the concept lacks (busy, conflict, unavailable model, video) | Global Constraints: `#tag-generation-status` text only |
@@ -1914,8 +1969,9 @@ targets of the longest legal path (1024 code points, even escaped as `\uXXXX`) f
   Auto-tagging." (the server's own sentence), so no job starts only to fail.
   Start validates the fields as Generate does (native validity; nothing is sent).
   The concept comparison (`images-concept.spec.mjs`) uses the concept page itself as
-  the oracle, with Add to collection and Remove links… removed and a copy of its Add
-  tags button reading "Auto-tag selected" inserted after it.
+  the oracle, with Remove links… removed and a copy of its Add tags button reading
+  "Auto-tag selected" inserted after it (Add to collection compares in place since
+  #331, D138).
 - **D90** (plan D67, ruling T3-bulk): the dialog's **How to apply** radio group
   (`.field[role=radiogroup]`, radios beside their labels): **Apply automatically**
   (checked; primary "Start auto-tagging") starts the `image-tags` job
@@ -2006,7 +2062,7 @@ busy". Units:
 
 | Difference | Record |
 | --- | --- |
-| Add tags ships (Add to collection and Remove links… still omitted) | D8 (amended) |
+| Add tags ships (Remove links… still omitted; Add to collection ships with #331, D138) | D8 (amended) |
 | Auto-tag selected and its dialog | D85 (ruling L1) |
 | How to apply and the Review each image session | D90 (ruling T3-bulk) |
 | Suggestions count in the bulk dialog | D91 |
@@ -2113,13 +2169,242 @@ npm run --prefix tools/vendor test:workflows -- workspace-tagging-lifecycle.spec
   --project firefox --project webkit
 ```
 
+### Collections contract (#331)
+
+A collection is a named, hand-picked group of images (milestone 7 "Additional
+organization"). It never copies, moves or changes a file, a prompt, a tag, a category
+or a favorite. Legacy pages have no collections, so the legacy coverage table gains no
+row.
+
+**Database** (`database/image_collections.py`, mixed into `PromptDatabase`):
+- `image_collections(id, name, name_key UNIQUE, created_at)` and
+  `image_collection_items(collection_id → image_collections ON DELETE CASCADE,
+  root_key, relative_path, added_at, PRIMARY KEY (collection_id, root_key,
+  relative_path))`, plus `idx_image_collection_items_file(root_key, relative_path)`.
+  `migrate_image_collections` runs right after the saved-views migration inside its own
+  savepoint, creates the three objects empty with no backfill, changes no existing
+  table and reads only `sqlite_master` and its own tables, so its cost does not depend
+  on library size (checked with an SQLite authorizer on a 23,000-image library; a
+  second run is a no-op). Startup validates structure only; restore runs the full value
+  check.
+- A name is stored as typed after normalization: trimmed, each run of whitespace one
+  space, 1 to 100 characters, no control characters (Unicode `Cc`). `name_key` is
+  `NFC(casefold(NFC(name)))`, so names are unique ignoring case and Unicode form (a
+  macOS NFD name matches the Windows NFC one). Restore checks `name_key` only as
+  non-empty bounded text and does not recompute it: Python's Unicode case tables differ
+  between versions, so a backup made on another Python must still restore.
+- A member names a file by location, exactly as a favorite does: the source folder's
+  real path (`_favorite_root_key`) plus the path below it (`_favorite_relative_key`),
+  case-folded on Windows, each at most 4,096 characters. A case-only rename keeps a
+  Windows member; two Linux files that differ only by case are two members; on macOS a
+  case-only rename hides the member, as it hides a favorite. The Windows folding is
+  tested on every OS through `ntpath`.
+- Each write is one `BEGIN IMMEDIATE` transaction, so two tabs adding the same new
+  name create one collection. A collection is created only when at least one file is
+  inserted.
+
+**Missing members (MC2, MC3):** a member whose file is moved, renamed or deleted
+outside PromptManager, or whose source folder is offline, is hidden and its row kept;
+it shows again when a file appears at the same path (which may be a different image,
+as with favorites). Nothing follows a moved file, and there is no "forget missing"
+cleanup; deleting the collection removes those rows.
+
+**Duplicate cleanup (MC4):** System › Maintenance duplicate files moves the removed
+copy's memberships to the kept copy as a union, in the transaction that moves its
+favorite and image tags, and the job log reads "Moved N collection membership(s) to the
+kept copies". The legacy gallery page's duplicate delete does not move them (as for
+favorites).
+
+**Restore (MC8):** `collections` joins the restore stage's summary keys and
+`image_collections`'s `MAX(rowid)` its fingerprint; membership edits are not
+fingerprinted, as favorite and tag edits are not. A backup made before collections
+restores with no collections (the tables are created empty); the pre-restore safety
+backup keeps the old ones. The restore review shows no collections count.
+
+**Routes** (`py/api/collections.py`; every answer `{"success": bool, ...}`):
+
+| Route | Request | Success |
+| --- | --- | --- |
+| `GET /prompt_manager/collections` | none | `{collections: [{id, name, count, missing}]}` in name-key, then id, order |
+| `POST /prompt_manager/collections/items` | `{collection_id}` or `{name}`, and `items` | `{collection: {id, name}, created, added, already, not_found}` |
+| `DELETE /prompt_manager/collections/items` | `{collection_id, items}` | `{collection: {id, name}, removed}` |
+| `GET /prompt_manager/collections/membership` | query `root_index`, `root_scope`, `path`, each once | `{collection_ids: [...]}` |
+| `PATCH /prompt_manager/collections/{id}` | `{name}` | `{collection: {id, name}}` |
+| `DELETE /prompt_manager/collections/{id}` | none | `{deleted: true}` |
+
+- Ids are `[1-9][0-9]{0,8}` (path) or that int (body); anything else, or a body with
+  both or neither of `collection_id` and `name`, is 400.
+- `items` holds 1 to 500 targets `{root_index, root_scope, path}`, each checked by the
+  shared `py/api/media_target.py` (the image-annotation routes use the same check): 400
+  for a malformed shape, 403 for a traversal, drive, backslash or non-media path. A
+  `root_scope` that no longer names its root is 409 `source-changed` ("The sources
+  changed. Refresh the images and try again."). Any invalid item refuses the request
+  before a write; two targets that fold to one key count once.
+- Add keeps only files the source's listing holds and counts the rest in `not_found`;
+  a new name with nothing inserted is 404 `files-not-found` and creates nothing. A name
+  matching an existing collection ignoring case adds to it. Remove accepts any
+  well-formed target, listed or not.
+- An unknown id is 404 `collection-not-found`; a rename to another collection's name is
+  409 `name-taken` (a case variant of its own name is allowed). Any other failure is a
+  logged 500 with "Collections could not be read." or "The collection could not be
+  changed.", never the exception text.
+- Counts: `count` is the members the current listing of every configured source shows
+  (a folder configured twice counts once), `missing` the stored rows beyond it,
+  clamped at 0. Both come from one read snapshot of every member and stored count, so a
+  write between two reads cannot skew them. When any source's scan or read fails, both
+  are `null` for every collection and the names still return.
+- Every database read and every filesystem call (realpath, the scans) runs in the
+  executor; nothing touches the filesystem on the event loop.
+- The Images listing (`/prompt_manager/images/output?mode=workspace`) accepts
+  `collection` (`""` or an id) among `LISTING_FILTER_KEYS`; the cursor's scope hash
+  includes it, so a cursor from one collection is refused on another. An unknown
+  collection is 404 `collection-not-found`, checked before any per-source read; a
+  failed members read is the listing's 500, never an unfiltered page. A collection and
+  `favorites=1` are never both applied (the collection wins).
+
+**Limits (MC12):** names 1 to 100 characters; at most 500 files per add or remove; no
+limit on the number of collections or members.
+
+**Selection bar and "all N matching" (#331 PR 3 ruling):** the collection routes take
+explicit files only (no listing filter). With "All N matching selected" (D135), Add to
+collection and Remove from collection send the loaded images only when every match is
+loaded and checked and there are at most 500; otherwise they toast "Select at most
+500 images." and open nothing. Over 500 explicitly checked files toast the same.
+
+Recorded deviations:
+- **D136** (amends D28): the concept's COLLECTIONS group (`#collections-nav`,
+  index.html 28–31) without the "PROPOSED" badge. One `links()` row per collection
+  (folder icon, name, `.nav-count`) in name-key order, linking `#/collection/<id>`. The
+  count follows the nav-count rules of D51 (hidden until loaded, absent when it could
+  not be read, never a placeholder) and is re-read 300 ms after the last confirmed add,
+  remove, rename or delete; the list is re-read after a restore. The label is hidden
+  while no collection exists. At 950 px and below the concept hides
+  `#collections-nav`, which would leave a phone no way to open a collection: the rows
+  show in the wrapped nav (`workspace.css`, `/* D136 */`), the label stays hidden as
+  every sidebar label is, and the counts are hidden as on every concept nav row at that
+  width. (Open question for the maintainer: CUJ 15 asks for "with counts" at phone
+  width; production follows the concept's phone rule and hides them.) A failed list
+  read keeps the last rows.
+- **D137**: the collection page `#/collection/<id>` is Images filtered to the
+  collection: the collection's name is the h1, "Collection" the breadcrumb and title
+  (concept a1186–1189), with the concept's Images description and every Images control.
+  Its heading leads with **Edit** ("Edit collection", "Collection name", Delete…
+  (danger), Cancel, Save; the saved views' D51 dialog with collection nouns): "Collection
+  renamed.", a taken name stays in the dialog as "A collection named “X” already
+  exists.", "The collection could not be renamed. Try again." keeps it open. Delete…
+  asks "Delete “X”? Its N images stay where they are; only the collection is removed."
+  (N the stored count) and toasts "Collection deleted." (a 404 counts as deleted),
+  then opens `#/images`; a failure toasts "The collection could not be deleted. Try
+  again.". A `p.small` after the description reads "N image(s) in this collection
+  is/are not in your current sources." when members are missing. An unknown id reads
+  "This collection no longer exists." with an "Open Images" link; a failed list reads
+  "Collections could not be loaded. Your saved data has not changed." with Retry. The
+  page neither restores nor replaces the remembered Images state, and its favorites
+  filter is off (DC10).
+- **D138** (amends D8): **Add to collection** follows Set category and opens the
+  concept's `simpleInput` ("Add selected images to collection", "Collection name")
+  with the action "Add" (the concept's "Apply in prototype"), an empty field (the
+  concept pre-fills its sample "Experiments") of at most 100 characters, and a
+  `datalist` of the existing names (`#action-input-suggestions`; the field is then a
+  combobox, and Chromium draws its suggestion arrow inside the field, which the
+  geometry comparison does not see). An empty field keeps the dialog on it. The toasts: "Added N image(s) to
+  “X”." plus " N was/were already in it." and " N could not be found.", "All selected
+  images were already in “X”.", "None of the selected images could be found. Refresh
+  the images and try again." (404 `files-not-found`), "The sources changed. Refresh
+  the images and try again." (409; the sources and the listing are read again, which
+  clears the selection), "That collection no longer exists." and "The images could not
+  be added. Try again.". Otherwise the selection stays and the listing is not reloaded.
+  **Remove from collection** (beyond the concept) follows Add while a collection filter
+  is active (a collection page or a saved view on one), with no confirmation: "Removed
+  N image(s) from “X”.", then the loaded range is read again in place, keeping the
+  checked files still listed. "That collection no longer exists." (404, a collection
+  deleted meanwhile) reads the listing again, which then shows the deleted-collection
+  page; "The sources changed. Refresh the images and try again." (409) reads the
+  sources and listing again and clears the selection; "The images could not be removed.
+  Try again." keeps everything. A second press while it runs is ignored.
+- **D139** (beyond the concept): the shared inspector's footer ends with
+  **Collections…** (the folder icon) after Prompt history, for a listed file only
+  (`rooted-file` with its scope), in the Images side panel and the viewer's details;
+  Inspect a file and the Prompt library's galleries have no client store and no
+  button. Its dialog "Collections": "Loading collections…" while the file's membership
+  is read ("Collections could not be loaded." with Retry on failure; Save waits), one
+  concept checkbox row (`label.tag-suggestion`) per collection in sidebar order,
+  checked where the file is a member ("No collections yet. Name one below." without
+  any), a "New collection" field (at most 100 characters; a name matching an existing
+  collection adds to it), then Cancel and Save. Focus starts on the field and moves to
+  the first checkbox once they are listed. A typed name equal to a listed collection
+  ignoring case is that collection's checkbox, checked (never an add then a remove); a
+  name that cannot be a collection name is refused on the field with "The collections
+  could not be updated. Try again.", nothing is sent and the checkbox edits stay. Save
+  sends one request per change: the new name, each newly checked collection, each
+  unchecked one; nothing changed sends nothing and closes. All sent: the dialog closes
+  with "Collections updated.". A failure keeps it open with "The collections could not
+  be updated. Try again." and the membership read again, Save waiting until that read
+  lands; × and Escape wait while it saves. The page hears of the changes that went
+  through only when the dialog closes (a failure included), so a reload cannot remove
+  the dialog before its message is read. On a collection page a change to the open
+  collection then reloads the listing at once from the side panel, and from the viewer
+  only when the viewer closes, so the viewer keeps its place (no listing request while
+  it is open). Closing returns focus to the trigger, or, when a re-render replaced it,
+  to the inspector's current Collections… button or its heading. The
+  inspector comparison draws the button into the concept's panel footer and full
+  details after Prompt history (a copy of its Prompt history button with the concept's
+  folder icon), so the browser lays out what follows.
+- **D140**: saved views' Images state gains `collection` (`""` or an id, the `source`
+  encoding); older rows read `""`, and a write with both `favorites: true` and a
+  collection is 400. Save view on a collection page stores the id; the view reopens
+  the collection with its other filters under the view's own name (D51), with Remove
+  from collection and without the missing note. A view whose collection was deleted
+  (MC6) lists nothing and reads "The collection this view uses was deleted."; it never
+  falls back to every image, and the view is not changed.
+
+| Difference | Record |
+| --- | --- |
+| COLLECTIONS group, rows, counts, phone width | D136 (amends D28) |
+| Collection page, Edit, missing note, deleted page | D137 |
+| Add to collection dialog and toasts; Remove from collection | D138 (amends D8) |
+| Inspector and viewer Collections… and its dialog | D139 |
+| Saved views keep a collection; MC6 empty state | D140 |
+
+Comparisons: `tests/browser/shell-concept.spec.mjs` compares `#collections-nav` with
+two seeded collections ("Landscapes" with 12 listed members, "Experiments" with 8), so
+the concept's rows and counts compare as they are, mapping out only the badge and, at
+390 px, the concept's hidden nav. `tests/browser/images-concept.spec.mjs` compares Add
+to collection in place in the selection bar, and the collection page's heading,
+breadcrumb, toolbar and filter row against the concept's `#collection-landscapes`
+(the grid is not compared: the concept's members come from categories).
+`tests/browser/collections-concept.spec.mjs` compares the bulk dialog with the
+concept's `simpleInput` (D138 maps "Apply in prototype" to "Add", the sample value to
+empty and adds the datalist), at 1600x1100 and 390x844 in dark and light, with a
+mutation check. `tests/browser/inspector-concept.spec.mjs` maps D139.
+
+Tests: Python `tests/test_image_collections_schema.py` (migration, validation,
+restore), `tests/test_image_collections.py` (operations, keys, the duplicate move),
+`tests/test_api_collections.py` (routes, rules, counts), `tests/test_api_workspace_images.py`
+(the `collection` filter), `tests/test_api_duplicate_files.py`, `tests/test_restore_stage.py`,
+`tests/test_saved_views.py`; units `tests/js/workspace-collections.test.js` (the client
+store and strings), `tests/js/workspace-images-selection.test.js` (the bar's order, the
+500 cap, the all-matching rule and every §7.4 string), `tests/js/workspace-file-inspector.test.js`
+(D139's button and dialog), `tests/js/workspace-toolbar.test.js` (the datalist);
+browser `tests/browser/workspace-collections.spec.mjs` (CUJ 1–15 on the fixture
+`tests/browser/workspace_collections_fixture.py`: twelve Landscapes members and one
+missing file, eight Experiments members, one loose file and a saved view on
+Experiments), `tests/browser/full-inspector.spec.mjs` (the footer with Collections…).
+
+```sh
+python3 -m unittest tests.test_image_collections_schema tests.test_image_collections tests.test_api_collections -q
+node --test tests/js/workspace-collections.test.js tests/js/workspace-images-selection.test.js tests/js/workspace-file-inspector.test.js
+npm run --prefix tools/vendor test:workflows -- workspace-collections.spec.mjs collections-concept.spec.mjs \
+  images-concept.spec.mjs inspector-concept.spec.mjs shell-concept.spec.mjs --project chromium --project touch
+```
+
 ### Prompt library cards and recorded deviations (#249 PR E)
 
 The shelf uses the shared concept toolbar and caption (E1) and the concept's
 `article.prompt-card` (app.js `renderLibrary` a586, style.css s653–708): `.card-art`
 (at most two previews), `.card-content` (`.card-title` with the title `button` and the
 category `span.chip`, the positive prompt block, `.card-meta`, `.tags`) and
-`.card-footer` (the rating and last-use line, then `.row-actions` with the positive
+`.card-footer` (the last-use line, then `.row-actions` with the positive
 `button.send.icon-btn` and "Explore history"). Rows stay windowed (`.library-row`).
 `tests/browser/workspace-library.spec.mjs` covers the markup, the clamp, the buttons and
 the measured row heights on Chromium and touch.
@@ -2129,8 +2414,8 @@ Recorded deviations on this page:
   "Used today", "Used yesterday", "Used N days ago" (under a week) or "Used <Mon D>"
   ("Used <Mon D>, <YYYY>" for a use in another calendar year, so last year's date never
   reads as this year's), counted in the viewer's local calendar days from `last_used_at` (a time ahead of the
-  browser reads as today), and "Not used yet" without one. Stars are `★` × rating with
-  a visually hidden "Rated N out of 5"; an unrated parent shows only the last-use words.
+  browser reads as today), and "Not used yet" without one. There are no rating stars
+  (D153, amending D23): every parent shows only the last-use words.
   The meta chip reads "− N negative variants" (singular for one). The title and Explore
   history are buttons that open `#/library/<id>`, as the concept's are; the trade-off
   against the old links is that they offer no open in a new tab, middle-click or URL on
@@ -2196,6 +2481,143 @@ npm run --prefix tools/vendor test:workflows -- library-concept.spec.mjs \
   workspace-library.spec.mjs --project chromium --project touch --project firefox --project webkit
 ```
 
+### Selecting prompts and Set category (#437)
+
+**D148** (beyond the concept, whose Library cards have no checkbox): the Prompt library
+selects prompts and sets their category in one step, the legacy admin page's bulk "Set
+Category". It is built only from the concept's Images parts:
+
+- Each `article.prompt-card` leads with the Images `.image-overlays > label > input`
+  checkbox ("Select <title>"), placed by the concept's overlay CSS over the art; the card
+  carries `position: relative` (the concept puts it on `.media-art`). The checkbox comes
+  before the title in tab order (on Images it follows the image button). Tab and
+  Shift+Tab across windowed cards reach it (the card's controls are `button,
+  input[type=checkbox]`); Home, End and the arrow keys still move between titles.
+- `div#library-bulkbar` sits between the toolbar and the caption, empty until something
+  is checked, then the concept's `.bulkbar`: "N selected", Set category, Clear (D151
+  adds Add tags before Set category, in the Images bar's order). The
+  selection is a set of prompt ids over the loaded shelf: windowed-out cards keep their
+  state, Load more keeps it, and any shelf reload (search, category, sort, tags, a
+  write) and leaving the page clear it. Clicking a checkbox toggles it and sets the
+  anchor; Shift+click applies the anchor's range with the box's new state (the D135
+  helpers). Escape in the bar and Clear clear it, focus to the heading. There is no
+  Shift+arrow, long-press or "Select all N matching" on the Library (D151 adds
+  Ctrl/Cmd+A and "Select all N loaded").
+- Set category opens the shared `simpleInput` dialog ("Categorize selected prompts",
+  `input#action-input` labelled Category, at most 100 characters, the Library's
+  categories as suggestions). A blank value keeps it open; a control character is
+  refused on the field ("A category cannot contain control characters."). More than
+  1,000 selected toasts "Select at most 1,000 prompts." and opens nothing.
+- Apply posts `{prompt_ids, category, expected_epoch}` to `POST
+  /prompt_manager/bulk/category`, the epoch being the shelf's first page's
+  `library_epoch` (kept with a returning visit; a later page from another epoch is a
+  shelf error). The server checks the epoch, then sets the category on the Library
+  parents among the ids in one `BEGIN IMMEDIATE` transaction, writing (and registering as
+  Library parents, as the editor's PATCH does) only rows whose category differs, and
+  answers `{library_epoch, category, updated, unchanged, not_found}`. A restored storage
+  answers 409 `storage-changed` and writes nothing. The legacy `{prompt_ids, category}`
+  and `{images_filter, category}` bodies are unchanged.
+- Outcomes: "Category “X” set on N prompts." (plus "M could not be found."), then the
+  shelf and its categories reload and cached Library views and the sidebar counts are
+  invalidated, also when the page was left during the write. None found: "None of the
+  selected prompts could be found. The Library was reloaded." Storage restored: "Prompt
+  storage was restored since this page loaded. No category was changed." and a reload.
+  No answer, 502, 503 or 504: "The category change could not be confirmed. The Library
+  was reloaded." Any other refusal: "Category could not be set. Try again.", the
+  selection kept. Every bar action (and D151's select-all offer) is `aria-disabled` while
+  a write runs.
+- The bar's DOM and write loop live in `library-selection.mjs`; the toast text
+  (`setOutcome` with the noun "prompt"), the range helpers, the count and the dialog are
+  the Images bar's. One bar core shared with `images-selection.mjs` is a follow-up.
+
+Tests: `tests/browser/workspace-library-selection.spec.mjs` (set, blank and control
+characters, Clear/Escape/filter, Shift+click range and Tab across the window edge, the
+fence after a return visit, an unseen restore, refused and unconfirmed writes);
+`tests/browser/library-concept.spec.mjs` ("the Library selection checkbox and bar match
+the concept's Images ones", 1600x1100 and 390x844, both themes, with a moved-checkbox
+mutation); units `tests/js/workspace-library-selection.test.js`,
+`tests/js/workspace-library-state.test.js` (the epoch); Python
+`tests/test_bulk_library_category.py`. A Python change: ComfyUI needs a restart.
+
+### Adding tags to selected prompts and selecting all (#442)
+
+**D151** (beyond the concept, on D148's bar): the Prompt library adds tags to the selected
+prompts in one step, the legacy admin page's bulk "Add Tags", and selects every loaded
+prompt, as its select-all checkbox did (#444 ruling). The concept has neither.
+
+- The bar reads "N selected", Add tags, Set category, Clear: the Images bar's order.
+- Select all of the loaded shelf: Ctrl/Cmd+A in the grid (Images' key check; the search
+  box keeps its text meaning) selects every loaded prompt, windowed-out cards included,
+  and Escape in the grid clears it with focus kept on the card. "Select all N loaded"
+  follows the count while some loaded prompts are unchecked and at most 1,000 are loaded;
+  it selects them all and focus moves to Add tags; N follows each loaded batch. It sits
+  where D135's "Select all N matching" sits, but that offer appears once every loaded
+  image is checked; this one never reaches beyond the loaded prompts. No "all matching".
+- Add tags opens the shared `simpleInput` dialog ("Tag selected prompts", "Tags,
+  separated by commas", the tag vocabulary `GET /prompt_manager/tags` as suggestions, the
+  list the editor offers). The field refuses no tag, more than 20, a tag over 50
+  characters or with a control character, each with its sentence. Apply posts
+  `{prompt_ids, tags, expected_epoch}` to `POST /prompt_manager/bulk/tags`: prompt tags
+  only, never image tags, auto-tags or pair tags. Names are normalized as image tags are
+  (trim, no commas, casefold repeats folded); registry rows are found and inserted by
+  their exact name and only when some prompt gets them; a prompt gets only the names it
+  lacks by casefold. A prompt whose tags would pass 20 is skipped (`over_limit`), so the
+  editor can still save it. The answer is `{library_epoch, tags, updated, unchanged,
+  not_found, over_limit}`; the toast "Added “a, b” to N prompts." adds already-had,
+  skipped and not-found sentences. Storage restored: "Prompt storage was restored since
+  this page loaded. No tags were added."; unconfirmed: "Adding the tags could not be
+  confirmed. The Library was reloaded."; refused: "Tags could not be added. Try again."
+  The legacy `{prompt_ids, tags}` body is unchanged; both forms refuse a body that is not
+  a JSON object with 400.
+
+Tests: `tests/browser/workspace-library-selection.spec.mjs` (Add tags with its suggestions,
+field refusals and restore refusal; Ctrl/Cmd+A and Select all N loaded across batches and
+during a write); `tests/browser/library-concept.spec.mjs` (the bar comparison now includes
+Add tags); units `tests/js/workspace-library-selection.test.js`; Python
+`tests/test_bulk_library_tags.py`. A Python change: ComfyUI needs a restart.
+
+### No 1–5 prompt rating in the workspace (#447)
+
+**D153** (maintainer ruling 2026-10-28: the star is only for favorites; the concept keeps
+a 1–5 rating): the workspace shows and edits no prompt rating. The concept's rating
+appears in four places, and production draws none of them:
+
+- Library cards: the concept's footer reads "★★★★★ · Used today"; production reads only
+  the last use (amends D23). `library-concept.spec.mjs` takes the stars and their " · "
+  out of the concept page itself before measuring (`dropConceptRatings`), so the footer
+  facts are compared at production's width; it first checks that every concept card had
+  them.
+- Sort: the concept's "Highest rated" is not offered on Images or the Library (amends
+  D14); the comparisons drop it from the concept's options. A saved view or a Library
+  visit that still names `sort: "rating"` opens on Recently added; the server keeps such a
+  view as it was stored.
+- Editor (New prompt, Edit recipe metadata, Create text revision): no `#recipe-rating`
+  "Rating" select; `library-editor-canonical-comparison.mjs` `mapConcept` drops the
+  concept's Rating field. The editor never sends `rating`. An edit sends only changed
+  fields, so a stored rating stays as it is; a new prompt and a text revision create an
+  unrated positive (the revised prompt keeps its own rating; the legacy admin page shows
+  the revision as unrated). A reuse that leaves metadata unapplied compares title,
+  category, tags and notes only. The edit token still covers `rating`, so a rating
+  changed on a legacy page while the editor is open is an ordinary metadata conflict
+  whose review shows no visible difference.
+- System > Statistics: no "Avg. rating" column and no "Rated images" total (amends D96);
+  `system-concept.spec.mjs` maps the concept's table without that column.
+
+Data and API are unchanged: the `rating` column (`CHECK` 1–5), `PUT
+/prompt_manager/prompts/{id}/rating`, `rating` in Library, Images and statistics
+answers, and the server's `rating` sorts. The legacy pages keep their ratings until the
+cutover (#334). Favorites (the image star) are unchanged.
+
+Tests: units `tests/js/workspace-library-editor-state.test.js` (a prompt rated 4 edited
+and saved keeps 4, no `rating` in any body, no rating field),
+`workspace-library-state.test.js`, `workspace-images-state.test.js`,
+`workspace-saved-views.test.js`, `workspace-system-overview.test.js`,
+`workspace-library-format.test.js`; browser `workspace-library-editor.spec.mjs` (the
+fixture's positive stored with rating 4 still reads 4 after Edit, a revision, Add
+negative and a reuse), `workspace-library.spec.mjs`, `workspace-images.spec.mjs`,
+`workspace-saved-views.spec.mjs`, `workspace-system.spec.mjs` and the concept
+comparisons above.
+
 ### Prompt history page and recorded deviations (#249 PR F-a)
 
 The history page (`#/library/<id>`) takes the concept's heading, summary, negative
@@ -3111,12 +3533,13 @@ the per-section hand-off panel, was retired by S12 (#329): every section is nati
   every check passed, some could not run, or issues need review. Close / Review missing
   links, which reviews a fresh Missing-file links plan (D100, S7); it never cleans up.
   Nothing is changed by a check.
-- **D96** (Statistics): "Usage statistics" without the sample note: the concept's
-  "Runs by category" table (Category / Runs / Avg. rating) with the real runs and
-  average rating of eligible parents per category ("Uncategorized", "Not rated"),
-  then a second table "Library totals" (Total / Count: Prompts, Tags, Images from the
-  legacy statistics, and Rated images: indexed images whose linked prompt is rated).
-  The Overview row says "Runs and ratings by category, with library totals." because
+- **D96** (Statistics; amended by D153): "Usage statistics" without the sample note: the
+  concept's "Runs by category" table without its Avg. rating column (Category / Runs)
+  with the real runs of eligible parents per category ("Uncategorized"), then a second
+  table "Library totals" (Total / Count: Prompts, Tags, Images from the legacy
+  statistics; no Rated images). The API still reports `average_rating` and
+  `rated_images`; the workspace does not show them.
+  The Overview row says "Runs by category, with library totals." because
   that is what the dialog shows (the concept's own dialog shows only "Runs by
   category"); per-prompt run counts, prompt tags and dated runs are recorded, but no
   table of them exists yet. A failed read toasts "Statistics could not
@@ -3346,7 +3769,11 @@ the per-section hand-off panel, was retired by S12 (#329): every section is nati
   mid-response is not discarded by the page: the server's 15-minute expiry removes it. When a restore the page saw running
   finishes, the cached Images and Library visits are dropped, the sidebar counts and
   saved views are read again and the open route is shown again (the Activity job view
-  stays). A refused file toasts "This file can't be restored: <validator reason>.". A
+  stays). #415: an open review (`openReviewDialog`) then closes as dismissed, since its plan no
+  longer describes storage; while any other dialog is open, all of that waits until the last
+  one has closed, so a dialog keeps what was typed and its next write meets the restored-storage
+  refusal, as when the page never saw the restore. `followRestore` (tests/browser/unseen-restore.mjs)
+  makes the page see a restore run and finish deterministically. A refused file toasts "This file can't be restored: <validator reason>.". A
   review value may wrap anywhere (`#modal-body > dl.meta-grid dd`, CSS under D105), so a
   long file name keeps the concept's two equal columns at phone width.
 - **D106** (Integrations, S10, Q8): the send bridge always asks for the ComfyUI window
@@ -4916,7 +5343,7 @@ category; without one it shows the category its link supplies.
   category can still be removed. An unexpected failure is a logged 500 ("The
   category could not be changed." / "Image categories could not be read.") without
   the exception text. The routes never read or write `prompts`, `prompt_pairs` or
-  `generated_images`; `POST /prompt_manager/bulk/category` is unchanged.
+  `generated_images`; `POST /prompt_manager/bulk/category` (prompts only) is not one of them.
 - Duplicate cleanup (System › Maintenance): the kept copy's own category wins. A
   kept copy without one receives the removed copy's (`COPY_SQL`, part of the
   cleanup's query-plan check). The removed copy's row is never deleted, so a
@@ -4934,7 +5361,8 @@ category; without one it shows the category its link supplies.
   the side panel, the viewer and History's media inspector (D145); Tags & categories
   counts images on each category row and lists image-only values (D143), and Manage
   category previews and renames image rows too (D144). Prompt categories are changed in
-  the Prompt library editor and on the legacy admin page's "Set Category". Concept
+  the Prompt library editor, the Prompt library's selection bar (D148, #437) and on the
+  legacy admin page's "Set Category". Concept
   comparisons, at 1600x1100 and 390x844 in both themes: `images-concept.spec.mjs` (the
   Set category dialog and the options), `inspector-concept.spec.mjs` (the chip),
   `history-concept.spec.mjs` (History's chip) and `vocabulary-visual.spec.mjs` (rows and
@@ -5572,8 +6000,26 @@ decide through the same `own_counts`.
 - A cached node counts at job start like the hook: a job failing later still counts it.
   A node that executes after its replay in the same job (RAM-pressure freed the entry)
   skips its own count but still consumes the pending first-use catch-up.
-- No backfill: earlier cached jobs keep their counts. The history page's "tracked runs"
-  can stay below the Library count for these prompts (#403).
+- No backfill: earlier cached jobs keep their counts.
+- History (#403): the receipt also keeps the raw, prepend and append text the node gave
+  the history capture. For a cached node whose graph inference leaves a field unknown,
+  the capture takes it from the receipt matching the cache entry
+  (`CachedRuns.recorded_fields`), as `inferred` / `graph_semantics`: the node did not run
+  in this job, the value follows from the run key (the whole upstream closure) plus the
+  cache proof. The job then joins the prompt's history scope, gets the
+  `positive_prompt_id` (and Pair) hint, and an image of it can be attributed through its
+  run, so the history page's "tracked runs" equals the Library count from the restart on.
+  Fail closed (fields stay unknown, as before): no matching receipt, a receipt that
+  disagrees with a field the graph knows, executions with equal outputs that recorded
+  different fields (a conflict that stays for that run key), a field over 64 KiB or
+  invalid Unicode, more than one run's history (1 MiB) in one receipt, or a full 8 MiB
+  field budget (least recently used receipts give up their fields first; their counting
+  stays). No new evidence vocabulary, so older builds still open the database.
+- Still different: jobs before the restart (no backfill); list batches (per item in the
+  Library, once per job in history); text with leading or trailing whitespace (the
+  Library stores it stripped, history matches the exact raw text, #438); a job in which the
+  node executes after its replay (RAM-pressure freed the entry: two samples, an
+  ambiguous capture); jobs while the history writer is fenced (restore, library move).
 - Tests: `tests/test_cached_runs.py` (whole jobs with the real nodes, a real database,
   fake host caches); `tests/comfyui/execution-text-link.spec.mjs` "#347" cases on the
   managed host: PreviewAny into each node kind, 3 jobs changing only the sampler seed,
@@ -5581,7 +6027,10 @@ decide through the same `own_counts`.
   modes, A, B, A from the owned `PMCycledText` under LRU, typed text, a downstream
   failure, a "run selected output" job that does not need the node, and a rewire to the
   negative input. The RAM-pressure eviction path cannot be forced on a real host and is
-  covered by the Python tests.
+  covered by the Python tests. Since #403 these cases also check the history page's
+  tracked runs equal the Library count, and `tests/test_execution_history_hooks.py`
+  (`CachedRuntimeHistoryTests`) runs the real history hook and writer with a real
+  `CachedRuns`.
 
 ```sh
 python3 -m unittest discover -s tests -p "test_cached_runs.py"
@@ -5680,10 +6129,12 @@ Context cases verify A,A,B as three jobs/two combinations/one positive snapshot
 when every positive field matches; all five registered PM variants; exact raw
 whitespace and Unicode; known empty negatives; Text negative roles and crossed
 pairs; fully cached repeats after trigger configuration changes; successful source
-text before downstream failure/interruption; unknown cached dynamic raw; unsupported
+text before downstream failure/interruption; cached dynamic raw from the reused execution (#403); unsupported
 transforms and duplicate list invocation ambiguity. Fresh values use
 `runtime_return`, cached effective STRING outputs use `host_cache_output`, and
-submitted literals use graph inference. SQL NULL means unknown; empty text is
+submitted literals use graph inference. A cached dynamic raw is unknown unless the
+#347 receipt of the execution the cache entry comes from proves it; it is then
+`inferred` / `graph_semantics` (graph identity plus a same-process cache receipt, #403). SQL NULL means unknown; empty text is
 known. Text is never trimmed or normalized by capture. Cache availability and
 inferred sampler roles prove neither sampler execution nor consumption, and list
 indices do not identify generated images. External trigger configuration changes
@@ -6037,9 +6488,11 @@ Deploying this backend needs a ComfyUI restart (Python and route changes).
 `#/tags` ("Tags & categories", page title "Tags") is a native workspace route; it no
 longer hands off to the legacy tag manager. That page stays reachable from Overview's
 "Open existing tag manager" card for tag search and sort, the untagged entry and
-multi-tag AND/OR browsing; bulk prompt actions (delete, tag, category) stay on the
-legacy admin page. A prompt's category is changed in the Prompt library editor or with
-the legacy admin page's "Set Category" (`POST /prompt_manager/bulk/category`, unchanged);
+multi-tag AND/OR browsing; bulk prompt delete stays on the legacy admin page (bulk tags:
+the Prompt library's selection bar, D151). A
+prompt's category is changed in the Prompt library editor, on many prompts at once with
+the Prompt library's selection bar (D148, #437), or with the legacy admin page's "Set
+Category" (both `POST /prompt_manager/bulk/category`);
 the Images selection bar's Set category sets images' own categories only (D140, #332),
 and Manage category renames a value on prompts, recipes and images together (D144). The screen consumes only the `/prompt_manager/vocabulary/*` routes
 above and adds no route, so PR B needs no ComfyUI restart: a browser refresh is enough.

database/library_commands.py

@@ -21,6 +21,10 @@ _UUID = re.compile(r"[0-9a-f]{8}(?:-[0-9a-f]{4}){3}-[0-9a-f]{12}")
 _TOKEN = re.compile(r"[0-9a-f]{64}")
 # Parent metadata stored on the prompt row; title lives in prompt_library_state.
 _PROMPT_COLUMNS = ("category", "rating", "notes")
+# Ids per IN (...) read: well under SQLite's oldest bound-parameter limit (999).
+_BULK_CHUNK = 500
+# The tags one prompt may carry: what the editor's save accepts (``validate_tags``).
+MAX_LIBRARY_TAGS = 20
 
 
 class LibraryCommandError(Exception):
@@ -85,6 +89,13 @@ def _validate_epoch(expected_epoch: str) -> None:
         raise ValueError("expected_epoch must be a canonical UUID")
 
 
+def _validate_bulk_ids(prompt_ids: list[int]) -> None:
+    if not isinstance(prompt_ids, list) or not prompt_ids:
+        raise ValueError("prompt_ids must be a non-empty list")
+    if any(type(pid) is not int or not 1 <= pid <= 2**63 - 1 for pid in prompt_ids):
+        raise ValueError("prompt_ids must be positive integers")
+
+
 def _validate_guards(positive_id: int, expected_epoch: str, if_match: str) -> None:
     if type(positive_id) is not int or not 1 <= positive_id <= 2**63 - 1:
         raise ValueError("Invalid positive prompt ID")
@@ -308,3 +319,134 @@ class LibraryCommandOperations:
                 positive_created=False,
                 pair_created=pair_created,
             )
+
+    def bulk_set_library_category(
+        self, prompt_ids: list[int], category: str, *, expected_epoch: str
+    ) -> dict:
+        """Set one category on many Library parents, all or nothing (#437).
+
+        Guarded by the storage epoch the page read. Only rows whose category
+        differs are written (their ``updated_at`` and edit token change) and
+        registered as Library parents, as the editor's PATCH does; ids
+        that are missing or are not Library parents count as not found.
+        """
+        _validate_epoch(expected_epoch)
+        _validate_bulk_ids(prompt_ids)
+        prompt_ids = list(dict.fromkeys(prompt_ids))
+        if not isinstance(category, str) or not category.strip():
+            raise ValueError("Category is required")
+        validate_category(category)
+        category = category.strip()
+        with self._pair_transaction() as conn:
+            epoch = self._library_epoch(conn, expected_epoch)
+            current = {}
+            for start in range(0, len(prompt_ids), _BULK_CHUNK):
+                chunk = prompt_ids[start : start + _BULK_CHUNK]
+                marks = ",".join("?" for _ in chunk)
+                current.update(
+                    conn.execute(
+                        f"SELECT p.id,p.category FROM prompts p "  # nosec B608
+                        f"WHERE p.id IN ({marks}) AND {_ELIGIBLE}",
+                        chunk,
+                    ).fetchall()
+                )
+            changed = [pid for pid, value in current.items() if value != category]
+            for pid in changed:
+                self._ensure_library_parent(conn, pid)
+            now = utc_now_iso()
+            conn.executemany(
+                "UPDATE prompts SET category=?,updated_at=? WHERE id=?",
+                [(category, now, pid) for pid in changed],
+            )
+            return {
+                "library_epoch": epoch,
+                "category": category,
+                "updated": len(changed),
+                "unchanged": len(current) - len(changed),
+                "not_found": len(prompt_ids) - len(current),
+            }
+
+    def bulk_add_library_tags(
+        self, prompt_ids: list[int], tags: list[str], *, expected_epoch: str
+    ) -> dict:
+        """Add tags to many Library parents, all or nothing (#442).
+
+        Names are normalized once (trim, the vocabulary name rules, casefold
+        repeats folded, at most 20). A parent is given only the names it lacks,
+        compared with ``casefold`` against its own tags; registry rows are
+        found and inserted by their exact name, and only for names some parent
+        gets. A parent whose tags would pass 20 is skipped (``over_limit``) so
+        the editor can still save it. Pair and image tags are not touched.
+        """
+        from .annotations import ImageTagLimitError, normalize_image_tags
+
+        _validate_epoch(expected_epoch)
+        _validate_bulk_ids(prompt_ids)
+        prompt_ids = list(dict.fromkeys(prompt_ids))
+        too_many = f"At most {MAX_LIBRARY_TAGS} tags can be added at once"
+        try:
+            tags = normalize_image_tags(tags)
+        except ImageTagLimitError:
+            # Its message names images; this is a prompt route.
+            raise ValueError(too_many) from None
+        if not tags:
+            raise ValueError("tags must name at least one tag")
+        if len(tags) > MAX_LIBRARY_TAGS:
+            raise ValueError(too_many)
+        with self._pair_transaction() as conn:
+            epoch = self._library_epoch(conn, expected_epoch)
+            held = self._library_parent_tags(conn, prompt_ids)
+            additions, over_limit = {}, 0
+            for pid, names in held.items():
+                present = {name.casefold() for name in names}
+                missing = [name for name in tags if name.casefold() not in present]
+                if not missing:
+                    continue
+                if len(names) + len(missing) > MAX_LIBRARY_TAGS:
+                    over_limit += 1
+                    continue
+                additions[pid] = missing
+            wanted = list(
+                dict.fromkeys(n for names in additions.values() for n in names)
+            )
+            tag_ids = self._ensure_tags(conn, wanted)
+            now = utc_now_iso()
+            for pid, names in additions.items():
+                conn.executemany(
+                    "INSERT OR IGNORE INTO prompt_tags (prompt_id, tag_id) "
+                    "VALUES (?, ?)",
+                    [(pid, tag_ids[name]) for name in names],
+                )
+                self._ensure_library_parent(conn, pid)
+                conn.execute("UPDATE prompts SET updated_at=? WHERE id=?", (now, pid))
+            return {
+                "library_epoch": epoch,
+                "tags": tags,
+                "updated": len(additions),
+                "unchanged": len(held) - len(additions) - over_limit,
+                "not_found": len(prompt_ids) - len(held),
+                "over_limit": over_limit,
+            }
+
+    def _library_parent_tags(
+        self, conn: sqlite3.Connection, prompt_ids: list[int]
+    ) -> dict[int, list[str]]:
+        """The Library parents among ``prompt_ids`` and their own tag names."""
+        held: dict[int, list[str]] = {}
+        for start in range(0, len(prompt_ids), _BULK_CHUNK):
+            chunk = prompt_ids[start : start + _BULK_CHUNK]
+            marks = ",".join("?" for _ in chunk)
+            for (pid,) in conn.execute(
+                f"SELECT p.id FROM prompts p WHERE p.id IN ({marks}) "  # nosec B608
+                f"AND {_ELIGIBLE}",
+                chunk,
+            ):
+                held[pid] = []
+            for pid, name in conn.execute(
+                "SELECT pt.prompt_id,t.name FROM prompt_tags pt "  # nosec B608
+                f"JOIN tags t ON t.id=pt.tag_id WHERE pt.prompt_id IN ({marks})",
+                chunk,
+            ):
+                if pid in held:
+                    held[pid].append(name)
+        return held

docs/application-redesign.md

@@ -12,7 +12,8 @@ code. This plan proposes implementation order and acceptance gates.
 - The native Prompt Library merged in [#210](https://github.com/ComfyAssets/ComfyUI_PromptManager/pull/210) for [#208](https://github.com/ComfyAssets/ComfyUI_PromptManager/issues/208): one positive parent, saved negative variants, full selected text and a shared filtered image gallery. It does not invent chronological runs from counters or scan dates.
 - [Durable history storage](prompt-history-storage.md) records job attempts and outcomes from [#212](https://github.com/ComfyAssets/ComfyUI_PromptManager/pull/212), frozen prompt contexts from [#213](https://github.com/ComfyAssets/ComfyUI_PromptManager/pull/213), and observed run-owned output locators from [#214](https://github.com/ComfyAssets/ComfyUI_PromptManager/pull/214). Those locators do not verify file bytes or establish gallery-image associations. That association contract and the chronological execution timeline remain pending under [#209](https://github.com/ComfyAssets/ComfyUI_PromptManager/issues/209).
 - [#215](https://github.com/ComfyAssets/ComfyUI_PromptManager/issues/215) brings the shell, Images and natural-height Prompt Library presentation into the approved geometry and themes while retaining their existing operations. [#216](https://github.com/ComfyAssets/ComfyUI_PromptManager/issues/216) adds the shared current-file inspector with exact generation facts, scoped provenance, raw/download evidence and loaded-collection viewer controls, including a slideshow that pauses at native video. This delivery does not establish whole-application parity, change the default entry point or constitute a release. Recorded chronological run/context browsing continues under [#209](https://github.com/ComfyAssets/ComfyUI_PromptManager/issues/209); image tagging/review, System consolidation and remaining organization capabilities continue under #196.
-- Local Inspect (#249, PRs L1–L2) replaces the `#/inspect` handoff with the concept's native "Inspect a file" page: one local image or video previews inline and opens the shared viewer and full inspector on a record read entirely in the browser, with nothing uploaded, imported or stored. The legacy metadata page (`web/metadata.html`, served at `/prompt_manager/gallery.html`) stays reachable from Overview; its capabilities map to the shared inspector, except its one-click COPY ALL summary, whose ruled replacement ("Copy settings summary") is pending. Recorded deviations D60–D66 are in [TESTING.md](../TESTING.md). Local roles equal to the server's resolver follow in PR L3.
+- Local Inspect (#249, PRs L1–L3) replaces the `#/inspect` handoff with the concept's native "Inspect a file" page: one local image or video previews inline and opens the shared viewer and full inspector on a record read entirely in the browser, with nothing uploaded, imported or stored. The legacy metadata page (`web/metadata.html`, served at `/prompt_manager/gallery.html`) stays reachable from Overview; its capabilities map to the shared inspector, including its one-click COPY ALL summary, which is the shared inspector's "Copy settings summary" (D117, LI-4). PR L3 ([#286](https://github.com/ComfyAssets/ComfyUI_PromptManager/pull/286)) shipped that button and local roles equal to the server's resolver (`web/js/workspace/local-resolver.mjs`, TESTING.md "Local roles (#249 PR L3, decision LI-1)"). Recorded deviations D60–D66 and D117 are in [TESTING.md](../TESTING.md).
+- Collections ([#331](https://github.com/ComfyAssets/ComfyUI_PromptManager/issues/331), three PRs) adds named, hand-picked groups of images, stored in the library database by file location as favorites are: a COLLECTIONS sidebar group whose rows open Images filtered to one collection (`#/collection/<id>`) with rename and delete, Add to collection and Remove from collection in the Images selection bar, a Collections… checklist in the shared inspector for the side panel and the viewer, and saved views that keep a collection. Membership survives restarts, source reordering and System › Maintenance duplicate cleanup; no file is copied, moved or changed.
 
 ## Scope and compatibility
 

prompt_manager.py

@@ -275,7 +275,12 @@ class PromptManager(PromptManagerBase, ComfyNodeABC):
         record_runtime_context(
             type(self).__name__, text, prepend_text, append_text, encoding_text
         )
-        self._record_cached_item(unique_id, (encoding_text,), storage_text)
+        self._record_cached_item(
+            unique_id,
+            (encoding_text,),
+            storage_text,
+            fields=((text, prepend_text, append_text),),
+        )
         return result
 
     @classmethod

prompt_manager_base.py

@@ -243,14 +243,24 @@ class PromptManagerBase:
             return False
 
     def _record_cached_item(
-        self, unique_id: Any, outputs: tuple, text: Any, negative_text: Any = None
+        self,
+        unique_id: Any,
+        outputs: tuple,
+        text: Any,
+        negative_text: Any = None,
+        fields: Optional[tuple] = None,
     ) -> None:
-        """Remember this execution's outputs and stored text for cached jobs (#347)."""
+        """Remember this execution's outputs and stored text for cached jobs (#347).
+
+        ``fields`` are the ``(raw, prepend, append)`` texts this execution gave
+        ``record_runtime_context`` for each output, so a cached job's history
+        can record them too (#403).
+        """
         if unique_id is None:
             return
         try:
             self._cached_runs().record_item(
-                unique_id, outputs, stored_facts(text, negative_text)
+                unique_id, outputs, stored_facts(text, negative_text), fields
             )
         except Exception as error:
             self.logger.debug(f"Could not record the cached-run receipt: {error}")

prompt_manager_pair.py

@@ -198,7 +198,11 @@ class PromptManagerPair(PromptManagerFocused):
             type(self).__name__, negative_text, "", "", negative_text, role="negative"
         )
         self._record_cached_item(
-            unique_id, (positive, negative_text), text, negative_text
+            unique_id,
+            (positive, negative_text),
+            text,
+            negative_text,
+            fields=((text, prepend_text, append_text), (negative_text, "", "")),
         )
         return result
 

prompt_manager_text.py

@@ -246,7 +246,12 @@ class PromptManagerText(PromptManagerBase, ComfyNodeABC):
         record_runtime_context(
             type(self).__name__, text, prepend_text, append_text, final_text
         )
-        self._record_cached_item(unique_id, (final_text,), storage_text)
+        self._record_cached_item(
+            unique_id,
+            (final_text,),
+            storage_text,
+            fields=((text, prepend_text, append_text),),
+        )
         return result
 
     @classmethod

py/api/prompts.py

@@ -8,8 +8,10 @@ import sqlite3
 from aiohttp import web
 
 try:
+    from ...database.library_commands import LibraryCommandError
     from ...database.pairs import PairDependencyError
 except ImportError:
+    from database.library_commands import LibraryCommandError
     from database.pairs import PairDependencyError
 from urllib.parse import quote
 
@@ -1128,7 +1130,14 @@ class PromptRoutesMixin:
     async def bulk_add_tags(self, request):
         """Bulk add tags to prompts."""
         try:
-            data = await request.json()
+            try:
+                data = await request.json()
+            except ValueError:
+                return bad_request("Invalid JSON body")
+            if not isinstance(data, dict):
+                return bad_request("Expected a JSON object")
+            if "expected_epoch" in data:
+                return await self._bulk_library_tags(data)
             prompt_ids = data.get("prompt_ids", [])
             new_tags = data.get("tags", [])
 
@@ -1180,6 +1189,8 @@ class PromptRoutesMixin:
                 )
             if "images_filter" in data:
                 return await self._bulk_category_by_filter(data)
+            if "expected_epoch" in data:
+                return await self._bulk_library_category(data)
             if not data.get("prompt_ids"):
                 return web.json_response(
                     {"success": False, "error": "No prompt IDs provided"}, status=400
@@ -1216,6 +1227,54 @@ class PromptRoutesMixin:
                 status=500,
             )
 
+    async def _bulk_library_tags(self, data: dict) -> web.Response:
+        """The workspace Library's fenced form (#442).
+
+        ``{prompt_ids, tags, expected_epoch}``: prompt tags only, never image tags.
+        """
+        return await self._fenced_library_write(
+            data, "tags", self.db.bulk_add_library_tags
+        )
+
+    async def _fenced_library_write(
+        self, data: dict, field: str, write
+    ) -> web.Response:
+        """One fenced Library bulk write: exact keys, checked ids, 409 on a restore."""
+        fields = {"prompt_ids", field, "expected_epoch"}
+        if set(data) != fields:
+            return bad_request(
+                "Request must contain exactly " + ", ".join(sorted(fields))
+            )
+        try:
+            prompt_ids = parse_bulk_prompt_ids(data["prompt_ids"])
+            result = await self._run_in_executor(
+                write, prompt_ids, data[field], expected_epoch=data["expected_epoch"]
+            )
+        except LibraryCommandError as exc:
+            return web.json_response(
+                {
+                    "success": False,
+                    "code": exc.code,
+                    "library_epoch": exc.epoch,
+                    "error": "Prompt storage was restored; refresh the library",
+                },
+                status=409,
+            )
+        except ValueError as exc:
+            return bad_request(str(exc))
+        return web.json_response({"success": True, **result})
+
+    async def _bulk_library_category(self, data: dict) -> web.Response:
+        """The workspace Library's fenced form (#437).
+
+        ``{prompt_ids, category, expected_epoch}``: a non-blank category on
+        the selected Library parents, refused with 409 ``storage-changed``
+        when prompt storage was restored since the page read its epoch.
+        """
+        return await self._fenced_library_write(
+            data, "category", self.db.bulk_set_library_category
+        )
+
     async def _bulk_category_by_filter(self, data: dict) -> web.Response:
         """Set the category on every linked positive a listing filter admits.
 

tests/browser/collections-concept.spec.mjs

@@ -0,0 +1,108 @@
+import {test,expect} from './fixtures.mjs';
+import {SIZES,THEMES,seedConceptAt,openConcept,openProduction,diffInventory,mutate} from './concept-inventory.mjs';
+
+// #331 D138: the Images selection bar's "Add to collection" dialog against the concept's own (`bulk-collection` →
+// simpleInput, app.js a1392–1398 and a1138), on the concept-shaped fixture, at 1600x1100 and 390x844 in both themes.
+// Paired screenshots are artifacts only; there are no PNG baselines.
+test.use({locale:'en-US',timezoneId:'UTC'});
+
+const CARDS='.gallery-body > .image-grid > .media-card';
+
+/**
+ * Runs in the page: the open "Add selected images to collection" dialog: its title, the body's tree, the field's
+ * label, input and suggestions, and the actions, with the concept's geometry and tokens.
+ */
+function measureAddDialog() {
+    const dialog=document.querySelector('dialog#modal[open]');
+    const css=(node,...keys)=>node?Object.fromEntries(keys.map(key=>[key,getComputedStyle(node)[key]])):null;
+    const name=node=>[node.localName,...[...node.classList].sort()].join('.');
+    const text=node=>node?.textContent.trim()??null;
+    const body=dialog.querySelector('#modal-body');
+    const field=body.querySelector(':scope > .field');
+    const label=field?.querySelector(':scope > label');
+    const input=field?.querySelector(':scope > input');
+    const list=input?.list??null;
+    const actions=body.querySelector(':scope > .dialog-actions');
+    const rect=dialog.getBoundingClientRect();
+    return {
+        title:text(dialog.querySelector('#modal-title')),
+        tree:[...body.children].map(name),
+        field:field&&{tree:[...field.children].map(name),height:Math.round(field.getBoundingClientRect().height),
+            ...css(field,'display','flexDirection','rowGap','marginBottom')},
+        label:label&&{text:text(label),for:label.htmlFor,...css(label,'fontSize','fontWeight','color')},
+        input:input&&{id:input.id,type:input.type,value:input.value,maxLength:input.maxLength,list:input.getAttribute('list'),
+            ...css(input,'height','paddingLeft','fontSize','color','backgroundColor','borderTopColor','borderTopLeftRadius')},
+        suggestions:list&&{id:list.id,display:getComputedStyle(list).display,options:[...list.options].map(option=>option.value)},
+        actions:{tree:[...actions.children].map(name),...css(actions,'display','justifyContent','columnGap','marginTop'),
+            buttons:[...actions.children].map(button=>({text:text(button),className:button.className,
+                ...css(button,'backgroundColor','color','fontSize','paddingTop','paddingLeft','borderTopColor')}))},
+        width:Math.round(rect.width),height:Math.round(rect.height),
+    };
+}
+
+/**
+ * D138: the action is "Add" (the concept's "Apply in prototype"), the field starts empty (no sample "Experiments") and
+ * holds at most 100 characters, and a hidden `datalist` of the existing names (the fixture's two) follows the field:
+ * the shared input dialog's suggestions list, as Set category's (#332 D140).
+ */
+function mapAddDialogConcept(reference) {
+    const [cancel,apply]=reference.actions.buttons;
+    const [field,...rest]=reference.tree;
+    return {...reference,tree:[field,'datalist',...rest],
+        input:{...reference.input,value:'',maxLength:100,list:'workspace-input-suggestions'},
+        suggestions:{id:'workspace-input-suggestions',display:'none',options:['Experiments','Landscapes']},
+        actions:{...reference.actions,buttons:[cancel,{...apply,text:'Add'}]}};
+}
+
+async function openAddDialog(page,main) {
+    await page.locator(CARDS).first().locator('.image-overlays > label > input[type="checkbox"]').check();
+    await page.locator(`${main} #bulkbar`).getByRole('button',{name:'Add to collection',exact:true}).click();
+    await expect(page.locator('dialog#modal[open] #modal-title')).toHaveText('Add selected images to collection');
+    await page.mouse.move(0,0);
+    return page.evaluate(measureAddDialog);
+}
+
+async function conceptAddDialog(page,theme) {
+    await openConcept(page,'images',theme);
+    return openAddDialog(page,'#main');
+}
+
+async function productionAddDialog(page,theme) {
+    await openProduction(page,'images',theme);
+    await expect(page.locator(CARDS).first()).toBeVisible();
+    return openAddDialog(page,'#workspace-main');
+}
+
+test.beforeEach(async({page,request},testInfo)=>{testInfo.manifest=await seedConceptAt(page,request);});
+
+for(const theme of THEMES)for(const [width,height] of SIZES) {
+    test(`add to collection dialog matches the concept ${theme} ${width}x${height}`,async({page},testInfo)=>{
+        await page.setViewportSize({width,height});
+        const concept=await conceptAddDialog(page,theme);
+        await page.screenshot({path:testInfo.outputPath(`concept-add-to-collection-${theme}-${width}.png`)});
+        const production=await productionAddDialog(page,theme);
+        await page.screenshot({path:testInfo.outputPath(`production-add-to-collection-${theme}-${width}.png`)});
+        expect(concept.input.value,'the concept pre-fills its sample').toBe('Experiments');
+        expect(diffInventory(mapAddDialogConcept(concept),production)).toEqual([]);
+    });
+}
+
+test('add to collection comparison detects mutations',async({page})=>{
+    await page.setViewportSize({width:1600,height:1100});
+    const concept=await conceptAddDialog(page,'dark');
+    const expected=mapAddDialogConcept(concept);
+    const production=await productionAddDialog(page,'dark');
+    expect(diffInventory(expected,production)).toEqual([]);
+    // Unmapped, the concept differs only where D138 is recorded.
+    expect(diffInventory(concept,production).map(item=>item.path).sort()).toEqual(expect.arrayContaining(
+        ['actions.buttons.1.text','tree.1','input.list','input.maxLength','input.value']));
+    const paths=async()=>diffInventory(expected,await page.evaluate(measureAddDialog)).map(item=>item.path);
+    await page.evaluate(()=>{document.querySelector('dialog#modal[open] label[for="action-input"]').textContent='Name';});
+    expect(await paths()).toEqual(['label.text']);
+    await page.evaluate(()=>{document.querySelector('dialog#modal[open] label[for="action-input"]').textContent='Collection name';});
+    const undoStyle=await mutate(page,{css:'dialog#modal #action-input { border-color: rgb(255, 0, 0); } dialog#modal .dialog-actions { gap: 31px; }'});
+    expect(await paths()).toEqual(expect.arrayContaining(['input.borderTopColor','actions.columnGap']));
+    await undoStyle();
+    await page.evaluate(()=>document.querySelector('#workspace-input-suggestions').remove());
+    expect(await paths()).toEqual(expect.arrayContaining(['suggestions']));
+});

tests/browser/concept-inventory.mjs

@@ -118,9 +118,27 @@ export async function openConcept(page,route,theme,motion='no-preference') {
     await expect(page.locator('#main #page-title')).toBeVisible();
 }
 
+/**
+ * D157 (#451): production opens Images in Masonry when the browser has no Default layout; the concept starts at Grid.
+ * A comparison (and a test of the Grid's own geometry) starts production from the concept's state, a stored Default
+ * layout of Grid, once per tab and only while this browser stored nothing (a test's own settings win).
+ */
+export async function startInGrid(page) {
+    if (page.conceptLayoutSeeded) return;
+    page.conceptLayoutSeeded = true;
+    await page.addInitScript(() => {
+        try {
+            if (!location.pathname.startsWith('/prompt_manager/') || sessionStorage.getItem('pm-d157-seeded')) return;
+            sessionStorage.setItem('pm-d157-seeded', '1');
+            if (localStorage.getItem('gallerySettings') === null) localStorage.setItem('gallerySettings', JSON.stringify({ workspaceLayout: 'grid' }));
+        } catch { /* A page without storage keeps the production default. */ }
+    });
+}
+
 /** Production at `#/<route>` with the D4 Theme select, once the sidebar's real counts (D27) have settled. */
 export async function openProduction(page,route,theme,motion='no-preference') {
     await page.emulateMedia({reducedMotion:motion});
+    await startInGrid(page);
     await page.goto('about:blank');await page.goto(`${PRODUCTION_URL}#/${route}`);
     await selectTheme(page,theme);
     await expect(page.locator('#workspace-main #page-title')).toBeVisible();
@@ -214,6 +232,41 @@ export async function withChipSource(page,root,origin,category=null) {
     };
 }
 
+/**
+ * #449 D154: the inspector's tag sections in the concept's prompt-section frame (workspace.css `.tag-section`). The concept
+ * has one bare `.tags` row after its prompt blocks (app.js a402), so the recorded rule is added and that row wrapped at
+ * runtime, the browser laying out what follows: `sections` lists `[className, label, name]` per section, the first
+ * holding the concept's row and each further one a copy of it. Returns the undo.
+ */
+export const D154_TAG_SECTION=`.tag-section { border-top: 1px solid var(--line); padding-top: 15px; margin-top: 15px; }
+.tag-section .role-label { color: var(--ca-muted); }
+.tag-section > .tags { margin-top: 0; }`;
+export const PROMPT_TAGS=[['linked-tags','Prompt tags','Prompt tags']];
+export async function withTagSections(page,root,sections=PROMPT_TAGS) {
+    const restoreRule=await mutate(page,{css:D154_TAG_SECTION});
+    const token=await page.evaluate(({root,sections})=>{
+        const container=document.querySelector(root);
+        const inner=container?.querySelector(':scope > .detail-inner')??container?.querySelector(':scope > div');
+        const row=inner&&[...inner.querySelectorAll(':scope > .tags')].find(node=>node.previousElementSibling?.matches('section.prompt-section'));
+        if(!row)throw new Error(`withTagSections: no tag row after the prompt blocks under ${root}`);
+        const anchor=row.nextSibling,parent=row.parentNode,added=[];
+        sections.forEach(([className,label,name],index)=>{
+            const section=document.createElement('section');section.className=`tag-section ${className}`;section.setAttribute('aria-label',name);
+            const heading=document.createElement('div');heading.className='prompt-heading';
+            const text=document.createElement('span');text.className='role-label';text.textContent=`${label} · ${row.children.length}`;
+            heading.append(text);section.append(heading,index?row.cloneNode(true):row);
+            parent.insertBefore(section,anchor);added.push(section);
+        });
+        const token=`tag-sections-${Math.random().toString(36).slice(2)}`;
+        (window.__tagSections??=new Map()).set(token,()=>{parent.insertBefore(row,added[0]);added.forEach(section=>section.remove());});
+        return token;
+    },{root,sections});
+    return async()=>{
+        await page.evaluate(token=>{window.__tagSections.get(token)();window.__tagSections.delete(token);},token);
+        await restoreRule();
+    };
+}
+
 /**
  * Inventory sections: {chrome, heading, toolbar, filterRow, bulkbar, caption, grid, cards, continuation, panel, library,
  * history, noPageScroll}. A section is null when its element is absent. The Images sections (toolbar … panel, #249 C7)
@@ -662,6 +715,13 @@ export function measureInspector(selector) {
             text:when(text,()=>({tag:text.tagName.toLowerCase(),className:classes(text),text:text.textContent,tabIndex:text.tabIndex,...font(text),...edges(text),
                 ...css(text,'whiteSpace','overflowWrap'),...box(text),lines:px(rect(text).height/parseFloat(getComputedStyle(text).lineHeight))}))};
     }
+    function tagSectionOf(section) {
+        const heading=section.querySelector(':scope > .prompt-heading'),label=heading?.querySelector('.role-label'),row=section.querySelector(':scope > .tags');
+        return {className:classes(section),name:section.getAttribute('aria-label'),...edges(section),...box(section),
+            heading:when(heading,()=>({...layout(heading),...edges(heading),...box(heading)})),
+            label:when(label,()=>({tag:label.tagName.toLowerCase(),text:label.textContent,...font(label),...box(label)})),
+            row:when(row,rowOf)};
+    }
     function factsOf(list) {
         const style=getComputedStyle(list);
         return {...css(list,'display','columnGap','rowGap','fontSize'),...edges(list),columns:style.gridTemplateColumns.split(/\s+/).length,...box(list),
@@ -695,6 +755,8 @@ export function measureInspector(selector) {
         badges:when(rows.find(row=>parts.indexOf(row)<firstSection),rowOf),
         sections:sections.map(sectionOf),
         tags:when(rows.find(row=>parts.indexOf(row)>firstSection),rowOf),
+        // #449 D154: the labelled tag sections after the prompt blocks, each with its frame, heading, label and chip row.
+        tagSections:parts.filter(item=>item.matches('section.tag-section')).map(tagSectionOf),
         facts:when(parts.find(item=>item.matches('dl.meta-grid')),factsOf),
         // Disclosures (viewer only) by their summary and frame; their content is the D20 production facts.
         details:parts.filter(item=>item.localName==='details').map(item=>{const summary=item.querySelector(':scope > summary');

tests/browser/fast_reset.py

@@ -28,6 +28,11 @@ import weakref
 from pathlib import Path
 from typing import Any, Callable, TypeVar
 
+try:
+    import fcntl
+except ImportError:  # Windows: no flock, and no /dev/shm to share
+    fcntl = None
+
 T = TypeVar("T")
 
 # Set to 1 to run seeders on the live database as before #384 (the equivalence
@@ -43,8 +48,13 @@ MEMORY_ROOT = Path("/dev/shm")
 # the browsers and the other workers; with less free than this, the scratch
 # copy goes on disk instead.
 MEMORY_FREE_BYTES = 32 * 1024 * 1024
-# Scratch folders are named <prefix><owner pid>-<random>.
+# Scratch folders are created as <prefix>new-<owner pid>-<random>, locked, then
+# renamed to <prefix>held-<owner pid>-<random>. The pid is for people only:
+# ownership is the flock on OWNER_LOCK, which every PID namespace sees (#419).
 SCRATCH_PREFIX = "pm-reset-"
+STAGING_PREFIX = f"{SCRATCH_PREFIX}new-"
+HELD_PREFIX = f"{SCRATCH_PREFIX}held-"
+OWNER_LOCK = "owner.lock"
 
 
 def _quote(name: str) -> str:
@@ -96,29 +106,40 @@ def table_digests(conn: sqlite3.Connection, schema: str = "main") -> dict[str, s
     return result
 
 
-def _process_alive(pid: int) -> bool:
+def _hold(folder: Path, *, create: bool) -> int:
+    """Take ``folder``'s owner lock without waiting; the open descriptor."""
+    flags = os.O_RDWR | (os.O_CREAT if create else 0)
+    fd = os.open(folder / OWNER_LOCK, flags, 0o600)
     try:
-        os.kill(pid, 0)
-    except ProcessLookupError:
-        return False
+        fcntl.flock(fd, fcntl.LOCK_EX | fcntl.LOCK_NB)
     except OSError:
-        return True  # it exists but belongs to someone else
-    return True
+        os.close(fd)
+        raise
+    return fd
 
 
 def remove_orphaned_scratch(root: Path) -> None:
     """Remove scratch folders of servers that died without cleaning up.
 
     A server killed outright (SIGKILL, a crash) never runs ``close``, and
-    tmpfs keeps its folder in memory until reboot. The folder name carries
-    the owner's process id; a folder whose process is gone is removed.
+    tmpfs keeps its folder in memory until reboot. A live owner holds the
+    flock on the folder's owner lock; the kernel drops it when the owner
+    dies, in whatever PID namespace it ran (a container with ``--ipc=host``
+    shares /dev/shm but not the pids, #419). A folder whose lock is free is
+    removed. Folders without a lock (staging, or named before #419) cannot
+    be proven orphaned and are left alone.
     """
-    if os.name != "posix":
-        return  # os.kill(pid, 0) would terminate the process on Windows
-    for folder in root.glob(f"{SCRATCH_PREFIX}*-*"):
-        pid = folder.name[len(SCRATCH_PREFIX) :].split("-", 1)[0]
-        if pid.isdigit() and not _process_alive(int(pid)):
+    if fcntl is None:
+        return
+    for folder in root.glob(f"{HELD_PREFIX}*"):
+        try:
+            fd = _hold(folder, create=False)
+        except OSError:
+            continue  # held by its owner, no lock to judge by, or gone
+        try:
             shutil.rmtree(folder, ignore_errors=True)
+        finally:
+            os.close(fd)
 
 
 class _SeedApi:
@@ -162,6 +183,7 @@ class ScratchSeeder:
         self._fallback = Path(directory)
         self._memory_root = memory_root
         self._directory: Path | None = None
+        self._owner_lock: int | None = None
         self._on_live = on_live
         self._scratch = None
         self._unsynced: "weakref.WeakSet[sqlite3.Connection]" = weakref.WeakSet()
@@ -266,6 +288,9 @@ class ScratchSeeder:
         if self._directory is not None:
             shutil.rmtree(self._directory, ignore_errors=True)
             self._directory = None
+        if self._owner_lock is not None:
+            os.close(self._owner_lock)  # after the removal: never a free lock
+            self._owner_lock = None
 
     def _scratch_directory(self) -> Path:
         """A folder in memory (tmpfs) where there is one, else ``directory``.
@@ -274,10 +299,35 @@ class ScratchSeeder:
         kernel is writing back the pages they rewrite.
         """
         root = self._memory_root
-        if root is not None and root.is_dir() and os.access(root, os.W_OK):
+        if (
+            fcntl is not None
+            and root is not None
+            and root.is_dir()
+            and os.access(root, os.W_OK)
+        ):
             remove_orphaned_scratch(root)
             if shutil.disk_usage(root).free >= MEMORY_FREE_BYTES:
-                prefix = f"{SCRATCH_PREFIX}{os.getpid()}-"
-                return Path(tempfile.mkdtemp(prefix=prefix, dir=root))
+                held = self._held_memory_directory(root)
+                if held is not None:
+                    return held
         self._fallback.mkdir(parents=True, exist_ok=True)
         return self._fallback
+
+    def _held_memory_directory(self, root: Path) -> Path | None:
+        """A new folder under ``root`` that this seeder holds the lock of.
+
+        It is locked under the staging name, which no cleaner judges, and
+        only then renamed: no cleaner ever sees it held-named and unlocked.
+        """
+        staging = Path(
+            tempfile.mkdtemp(prefix=f"{STAGING_PREFIX}{os.getpid()}-", dir=root)
+        )
+        try:
+            fd = _hold(staging, create=True)
+        except OSError:
+            shutil.rmtree(staging, ignore_errors=True)
+            return None  # no locks on this filesystem: use the disk
+        held = root / (HELD_PREFIX + staging.name[len(STAGING_PREFIX) :])
+        os.rename(staging, held)
+        self._owner_lock = fd
+        return held

tests/browser/full-inspector.spec.mjs

@@ -130,7 +130,7 @@ test('panel matches the concept order and copy',async({page,request})=>{
     // Linked, workflow inferred: every concept element.
     await expect(panel.locator('.panel-heading h2')).toHaveText(POSITIVE_TITLE);
     expect(await panelStructure(panel)).toEqual({outer:['img.detail-cover','div.detail-inner','div.detail-footer'],
-        inner:['div.panel-heading','p.small','div.tags','section.prompt-section.positive','section.prompt-section.negative','div.tags','dl.meta-grid']});
+        inner:['div.panel-heading','p.small','div.tags','section.prompt-section.positive','section.prompt-section.negative','section.tag-section.linked-tags','dl.meta-grid']});
     await expect(panel.locator('img.detail-cover')).toHaveAttribute('alt','');
     await expect(panel.locator('.panel-heading .chip')).toHaveText('PNG');
     await expect(panel.locator('.detail-inner > p.small')).toHaveText('complete.png');
@@ -139,16 +139,21 @@ test('panel matches the concept order and copy',async({page,request})=>{
     // #332 MC8: the chip names where its value came from, in visible text after the category.
     await expect(badges.locator('span.chip:not(.blue)')).toHaveText('Recipe category: Inspection · recipe');
     await expect(badges.locator('span.chip:not(.blue) .ca-visually-hidden')).toHaveText('Recipe category: ');
-    // Recipe tags are not the positive tags the Images Tags filter matches, so they stay plain chips (D21 ruling).
-    await expect(panel.locator('.detail-inner > .tags').last().locator('span.chip')).toHaveText(['fixture','geometric']);
-    await expect(panel.locator('.detail-inner > .tags button')).toHaveCount(0);
+    // Recipe tags are not the positive tags the Images Tags filter matches, so they stay plain chips (D21 ruling); D154
+    // (#449) labels their own section after the negative prompt, named for the recipe as D146's category chip is.
+    const recipeTags=panel.getByRole('region',{name:'Recipe tags',exact:true});
+    await expect(recipeTags).toHaveClass('tag-section linked-tags');
+    await expect(recipeTags.locator('.prompt-heading .role-label')).toHaveText('Recipe tags · 2');
+    await expect(recipeTags.locator('.tags > span.chip')).toHaveText(['fixture','geometric']);
+    await expect(panel.locator('.tag-section button')).toHaveCount(0);
     await expect(panel.locator('.meta-grid dt')).toHaveText(['Seed','Steps','CFG','Sampler','Scheduler','Model']);
     await expect(panel.locator('.meta-grid dd')).toHaveText(['9007199254740993','28','6.5','dpmpp_2m','karras','fixture-model.safetensors']);
     const footer=panel.locator('.detail-footer');
-    await expect(footer.locator('button')).toHaveText(['Send to ComfyUI','Prompt history']);
+    // #331 D139: every listed file's footer ends with Collections… (folder icon).
+    await expect(footer.locator('button')).toHaveText(['Send to ComfyUI','Prompt history','Collections…']);
     await expect(footer.locator('button.send')).toHaveAccessibleName(`Send prompt pair to ComfyUI: ${POSITIVE_TITLE}`);
-    await expect(footer.locator('button.send svg.icon')).toHaveCount(1);await expect(footer.locator('button:not(.send) svg.icon')).toHaveCount(1);
-    await expect(panel).not.toContainText(/prompt evidence|Recipe tags|Positive prompt tags/);
+    await expect(footer.locator('button.send svg.icon')).toHaveCount(1);await expect(footer.locator('button:not(.send) svg.icon')).toHaveCount(2);
+    await expect(panel).not.toContainText(/prompt evidence|Positive prompt tags/);
     // Execution recorded and embedded parameters (D21), with the extension chip naming an unverified format.
     reads.patch=data=>{for (const role of ['positive','negative']) data.prompts[role].source='execution';};
     await selectFile(page,'explicit-empty.png');await expect(panel.locator('.panel-heading h2')).toHaveText('explicit-empty.png');
@@ -157,27 +162,30 @@ test('panel matches the concept order and copy',async({page,request})=>{
     await selectFile(page,'ambiguous.png');await expect(panel.locator('.prompt-section.positive .prompt-text')).toHaveText('embedded positive');
     await expect(panel.locator('span.chip.blue')).toHaveText('Embedded parameters');
     await expect(panel.locator('.panel-heading .chip')).toHaveText('File extension PNG');await expect(panel.locator('.panel-heading .chip .ca-visually-hidden')).toHaveText('File extension ');
-    await expect(panel.locator('.detail-footer button')).toHaveText(['Send to ComfyUI']);
-    // Unlinked: no badge, category, tags, pair or history.
+    await expect(panel.locator('.detail-footer button')).toHaveText(['Send to ComfyUI','Collections…']);
+    // Unlinked: no badge, category, tags, pair or history; the footer holds only Collections… (D139).
     reads.patch=null;await selectFile(page,'empty.png');await expect(panel.locator('.panel-heading h2')).toHaveText('empty.png');
-    expect(await panelStructure(panel)).toEqual({outer:['img.detail-cover','div.detail-inner'],
+    await expect(panel.locator('.detail-footer button')).toHaveText(['Collections…']);
+    expect(await panelStructure(panel)).toEqual({outer:['img.detail-cover','div.detail-inner','div.detail-footer'],
         inner:['div.panel-heading','p.small','section.prompt-section.positive','section.prompt-section.negative','dl.meta-grid']});
     await expect(panel.locator('.meta-grid dd')).toHaveText(Array(6).fill('Not recorded'));
     // Ambiguous saved association: the same honest absence.
     reads.patch=data=>{data.library={...data.library,status:'ambiguous',positive_id:null,pair_id:null,title:null,category:null,tags:[],category_scope:null,tags_scope:null};};
     await selectFile(page,'workflow-only.png');await expect(panel.locator('.panel-heading h2')).toHaveText('workflow-only.png');
-    await expect(panel.locator('.chip:not(.blue)')).toHaveCount(1);await expect(panel.locator('.detail-inner > .tags button')).toHaveCount(0);
+    await expect(panel.locator('.chip:not(.blue)')).toHaveCount(1);await expect(panel.locator('.tag-section')).toHaveCount(0);
     await expect(panel.getByRole('button',{name:'Prompt history',exact:true})).toHaveCount(0);
     // D39 mixed scopes: each field follows its own source, so the prefix and the tag chips can disagree.
-    const tagRow=panel.locator('.detail-inner > .tags').last();
+    const tagRow=panel.locator('.tag-section.linked-tags > .tags');
     reads.patch=data=>{data.library={...data.library,status:'linked',category:'Subject',category_scope:'pair',tags_scope:'positive',tags:['alpha','beta']};};
     await selectFile(page,'complete.png');
     await expect(panel.locator('span.chip:not(.blue) .ca-visually-hidden')).toHaveText('Recipe category: ');
     await expect(tagRow.getByRole('button')).toHaveText(['alpha','beta']);
+    await expect(panel.locator('.tag-section.linked-tags .role-label')).toHaveText('Prompt tags · 2');
     reads.patch=data=>{data.library={...data.library,status:'linked',category:'Subject',category_scope:'positive',tags_scope:'pair',tags:['alpha','beta']};};
     await selectFile(page,'workflow-only.png');
     await expect(panel.locator('span.chip:not(.blue) .ca-visually-hidden')).toHaveText('Positive prompt category: ');
     await expect(tagRow.locator('span.chip')).toHaveText(['alpha','beta']);await expect(tagRow.getByRole('button')).toHaveCount(0);
+    await expect(panel.locator('.tag-section.linked-tags .role-label')).toHaveText('Recipe tags · 2');
     // Loading and error states render in `.detail-inner` as `.small` text with the existing buttons.
     reads.patch=null;await page.unroute('**/images/inspect/**');
     await page.route('**/images/inspect/**',route=>route.fulfill({status:500,json:{success:false,error:'broken'}}));
@@ -241,7 +249,7 @@ test('tag chip filters Images by that tag',async({page,request})=>{
     await page.goto('/prompt_manager/workspace#/images');await selectFile(page,'workflow-only.png');
     const panel=page.getByRole('complementary',{name:'Selected image details',exact:true});
     const listing=page.waitForRequest(item=>item.url().includes('/images/output?')&&new URL(item.url()).searchParams.has('tags'));
-    await panel.locator('.detail-inner > .tags').getByRole('button',{name:'geometric',exact:true}).click();
+    await panel.getByRole('region',{name:'Prompt tags',exact:true}).getByRole('button',{name:'geometric',exact:true}).click();
     expect(new URL((await listing).url()).searchParams.get('tags')).toBe('geometric');
     await expect(page.locator('article[data-image-key="0:inspector/workflow-only.png"]')).toBeVisible();
     await expect(page.locator('article[data-image-key="0:inspector/complete.png"]')).toBeVisible();
@@ -258,7 +266,7 @@ test('pair send and history appear only with real roles and link',async({page,re
     await expect(panel.getByRole('button',{name:'Send prompt pair to ComfyUI: explicit-empty.png',exact:true})).toBeVisible();
     await expect(panel.getByRole('button',{name:'Prompt history',exact:true})).toHaveCount(0);
     await selectFile(page,'empty.png');await expect(panel.locator('.panel-heading h2')).toHaveText('empty.png');
-    await expect(panel.locator('.detail-footer')).toHaveCount(0);await expect(panel.getByRole('button',{name:/^Send prompt pair/})).toHaveCount(0);
+    await expect(panel.locator('.detail-footer button')).toHaveText(['Collections…']);await expect(panel.getByRole('button',{name:/^Send prompt pair/})).toHaveCount(0);
 });
 
 test('no provenance, raw or download controls in the side panel',async({page,request})=>{
@@ -286,19 +294,19 @@ test('viewer details hold provenance, raw records and downloads',async({page,req
     await page.setViewportSize({width:1600,height:1100});const reads=await patchInspection(page);
     await fullFixture(page,request);let details=await viewerDetails(page,'complete.png');
     // The panel content without the cover, D117's Copy settings summary after the facts, then the two disclosures,
-    // then the pair send and history buttons (D20).
-    expect(await viewerStructure(details)).toEqual({outer:['div','button.send','button'],
-        inner:['div.panel-heading','p.small','div.tags','section.prompt-section.positive','section.prompt-section.negative','div.tags','dl.meta-grid','button.settings-summary','details','details']});
+    // then the pair send and history buttons (D20) and D139's Collections….
+    expect(await viewerStructure(details)).toEqual({outer:['div','button.send','button','button'],
+        inner:['div.panel-heading','p.small','div.tags','section.prompt-section.positive','section.prompt-section.negative','section.tag-section.linked-tags','dl.meta-grid','button.settings-summary','details','details']});
     await expect(details.locator(':scope > div > button')).toHaveText('Copy settings summary');
     await expect(details.locator(':scope > div > button.settings-summary')).toHaveCSS('white-space','normal');
     await expect(details.locator('.detail-cover,.detail-inner,.detail-footer')).toHaveCount(0);
     await expect(details.locator('.panel-heading h2')).toHaveText(POSITIVE_TITLE);await expect(details.locator('.panel-heading .chip')).toHaveText('PNG');
     await expect(details.locator(':scope > div > p.small')).toHaveText('complete.png');
     await expect(details.locator('span.chip.blue')).toHaveText('Workflow inferred');
-    await expect(details.locator(':scope > div > .tags').last().locator('span.chip')).toHaveText(['fixture','geometric']);
+    await expect(details.getByRole('region',{name:'Recipe tags',exact:true}).locator('.tags > span.chip')).toHaveText(['fixture','geometric']);
     await expect(details.locator('.meta-grid').first().locator('dd')).toHaveText(['9007199254740993','28','6.5','dpmpp_2m','karras','fixture-model.safetensors']);
     await expect(details.locator('summary')).toHaveText(['File & provenance','Workflow & raw metadata']);
-    await expect(details.locator(':scope > button')).toHaveText(['Send to ComfyUI','Prompt history']);
+    await expect(details.locator(':scope > button')).toHaveText(['Send to ComfyUI','Prompt history','Collections…']);
     await expect(details.locator(':scope > button.send')).toHaveAccessibleName(`Send prompt pair to ComfyUI: ${POSITIVE_TITLE}`);
     // File & provenance: the honest facts and the concept's "Prompt evidence" line; no per-fact source lines.
     const provenance=details.locator('details').first();await provenance.locator('summary').click();
@@ -309,7 +317,9 @@ test('viewer details hold provenance, raw records and downloads',async({page,req
     await expect(provenance.getByText('Prompt evidence: workflow inferred',{exact:true})).toBeVisible();
     await expect(provenance.getByText('Saved association; not verified against current file',{exact:true})).toBeVisible();
     await expect(provenance.getByRole('link',{name:'Download original',exact:true})).toBeVisible();
-    await expect(details).not.toContainText(/Configuration evidence|Embedded parameter evidence|prompt evidence|extension|Recipe tags/);
+    // D154 (#449) labels the recipe's tags in their own section; the other legacy labels stay gone.
+    await expect(details).not.toContainText(/Configuration evidence|Embedded parameter evidence|prompt evidence|extension/);
+    await expect(details.locator('.tag-section.linked-tags .role-label')).toHaveText('Recipe tags · 2');
     // Workflow & raw metadata: the concept's focusable `pre`, then the downloads with their unavailable reasons.
     const raw=details.locator('details').last();await raw.locator('summary').click();
     await expect(raw.locator('pre[tabindex="0"] > code').first()).toBeVisible();
@@ -323,9 +333,10 @@ test('viewer details hold provenance, raw records and downloads',async({page,req
     await expect(details.getByText('Prompt evidence: embedded parameters',{exact:true})).toBeVisible();
     await expect(details.locator('.panel-heading .chip')).toHaveText('File extension PNG');
     await expect(details.locator('details dl.meta-grid dd').nth(4)).toHaveText('PNG extension');await expect(details).not.toContainText('.PNG');
-    // No badge, no line; no resolved pair or link, no buttons after the details.
+    // No badge, no line; no resolved pair or link: after the details only D139's Collections….
     reads.patch=null;await page.keyboard.press('Escape');details=await viewerDetails(page,'empty.png');
-    expect((await viewerStructure(details)).outer).toEqual(['div']);
+    expect((await viewerStructure(details)).outer).toEqual(['div','button']);
+    await expect(details.locator(':scope > button')).toHaveText(['Collections…']);
     await details.getByText('File & provenance',{exact:true}).click();await expect(details).not.toContainText('Prompt evidence');
     // The history button leaves the viewer for the prompt's history.
     await page.keyboard.press('Escape');details=await viewerDetails(page,'complete.png');
@@ -464,7 +475,7 @@ test('full inspector keeps long literal content reachable on mobile',async({page
 
 test('embedded gallery uses the same full inspector',async({page,request})=>{
     await expect(await request.post('/__test__/reset',{data:{full_inspector:true}})).toBeOK();const list=await(await request.get('/prompt_manager/library')).json();const id=list.items.find(item=>item.text.startsWith('Inspector portrait')).id;
-    await page.goto(`/prompt_manager/workspace#/library/${id}`);await selectFile(page,'complete.png');const panel=inspector(page);await expect(panel.getByRole('heading',{name:POSITIVE_TITLE,exact:true})).toBeVisible();await expect(panel.locator('.meta-grid dt')).toHaveText(['Seed','Steps','CFG','Sampler','Scheduler','Model']);await expect(panel.locator('.detail-inner > .tags').last().locator('span.chip')).toHaveText(['fixture','geometric']);await expect(panel).not.toHaveAttribute('id','image-inspector');await openSelected(page);await expect(page.locator('#viewer-dialog .file-inspector')).toHaveCount(1);
+    await page.goto(`/prompt_manager/workspace#/library/${id}`);await selectFile(page,'complete.png');const panel=inspector(page);await expect(panel.getByRole('heading',{name:POSITIVE_TITLE,exact:true})).toBeVisible();await expect(panel.locator('.meta-grid dt')).toHaveText(['Seed','Steps','CFG','Sampler','Scheduler','Model']);await expect(panel.getByRole('region',{name:'Recipe tags',exact:true}).locator('.tags > span.chip')).toHaveText(['fixture','geometric']);await expect(panel).not.toHaveAttribute('id','image-inspector');await openSelected(page);await expect(page.locator('#viewer-dialog .file-inspector')).toHaveCount(1);
 });
 
 test('Prompt history to the shown route closes the viewer and re-renders it, as the concept go() does',async({page,request})=>{

tests/browser/images-concept.spec.mjs

@@ -30,9 +30,12 @@ function captionDay(seconds,now=new Date()) {
     return `${name} · ${date.toLocaleDateString('en-US',options)}`;
 }
 
-/** D8 (amended by #249 PR D): the selection bar's real actions only; Add to collection and Remove links… are left out. */
-const OMITTED_BULK=Object.freeze(['Add to collection','Remove links…']);
-const OMITTED_BULK_SELECTORS=Object.freeze(['bulk-collection','bulk-remove'].map(action=>`#bulkbar [data-action="${action}"]`));
+/**
+ * D8 (amended by #249 PR D and #331): the selection bar's real actions only; Remove links… is left out. Add to
+ * collection compares in place (D138).
+ */
+const OMITTED_BULK=Object.freeze(['Remove links…']);
+const OMITTED_BULK_SELECTORS=Object.freeze(['bulk-remove'].map(action=>`#bulkbar [data-action="${action}"]`));
 /** D85: Auto-tag selected follows the concept's Add tags, a concept button of its own text. */
 const AUTO_TAG_CLONE=Object.freeze([{selector:'#bulkbar [data-action="bulk-tags"]',text:'Auto-tag selected'}]);
 
@@ -89,7 +92,8 @@ export function mapImagesConcept(reference,{manifest,now=new Date(manifest.now*1
     const paths=new Map(manifest.images.map(image=>[image.filename,image.path]));
     // D14: the sort appends production's two sorts. Save view is the concept's own (saved views, #249 D16 retired).
     const toolbar={...reference.toolbar};
-    toolbar.items=toolbar.items.map(item=>item.id==='sort'?{...item,options:[...item.options,'Oldest first','Name Z–A']}:item);
+    // D153 (#447): production offers no "Highest rated".
+    toolbar.items=toolbar.items.map(item=>item.id==='sort'?{...item,options:[...item.options.filter(option=>option!=='Highest rated'),'Oldest first','Name Z–A']}:item);
     // D36: the category list is the Library's real list, sorted alphabetically (database/library.py `ORDER BY category
     // COLLATE NOCASE, category`); the concept lists its samples in prompt order (app.js 348). Only the order is mapped.
     // D141 (#332): the list also holds the listed images' own categories (`GET /prompt_manager/images/categories`),
@@ -464,7 +468,7 @@ for(const [width,height] of SIZES)for(const theme of THEMES)test(`images selecti
         'Add to collection','Export metadata','Remove links…','Clear']);
     expect.soft(diffInventory(stripExact(mapBulkbar(reference.bulkbar,0).bulkbar.bar),stripExact(removed.bar)),'bar without the omitted actions').toEqual([]);
     expect.soft(actual.bulkbar.bar.items.map(item=>item.text),'production bar').toEqual(['1 selected','Add tags','Auto-tag selected',
-        'Set category','Export metadata','Clear']);
+        'Set category','Add to collection','Export metadata','Clear']);
     expect.soft(compare(empty,emptyActual,manifest,['bulkbar','cards','panel']),'nothing checked').toEqual([]);
     expect.soft(compare(oracle,actual,manifest,['bulkbar','cards','panel']),'one card checked').toEqual([]);
     // The bar comparison notices a change to the production bar.
@@ -472,6 +476,48 @@ for(const [width,height] of SIZES)for(const theme of THEMES)test(`images selecti
     expect(compare(oracle,await inventoryOf(page,'#workspace-main'),manifest,['bulkbar']).map(item=>item.path)).toContain('bulkbar.bar.paddingTop');
 });
 
+/**
+ * #331 D137: the collection page's Edit leads the heading actions, drawn in the concept as a plain copy of its Import
+ * button reading "Edit" (no icon, no action), so the browser lays out what follows it; returns an undo.
+ */
+async function drawCollectionEdit(page) {
+    await page.evaluate(()=>{
+        const actions=document.querySelector('#main .heading-actions');
+        const edit=actions.querySelector('button').cloneNode(false);
+        for(const key of Object.keys(edit.dataset))delete edit.dataset[key];
+        edit.textContent='Edit';actions.prepend(edit);window.__collectionEdit=edit;
+    });
+    await settle(page);
+    return ()=>page.evaluate(()=>{window.__collectionEdit.remove();delete window.__collectionEdit;});
+}
+
+// #331 D137: `#/collection/<id>` against the concept's `#collection-landscapes` (app.js a1152–1157, a1186–1189): the
+// collection's name as the h1, "Collection" as the breadcrumb, the Images description, toolbar and filter row. The grid
+// is not compared: the concept's members come from its categories, production's from the stored collection.
+for(const [width,height] of SIZES)for(const theme of THEMES)test(`collection page heading, toolbar and filter row match the concept ${width}x${height} ${theme}`,async({page},testInfo)=>{
+    await page.setViewportSize({width,height});
+    const manifest=testInfo.manifest;
+    await openConcept(page,'collection-landscapes',theme);await page.mouse.move(0,0);await settle(page);
+    await expect(page.locator('#main #page-title')).toHaveText('Landscapes');
+    await expect(page.locator('#breadcrumb')).toHaveText('Collection');
+    const restore=await drawCollectionEdit(page);
+    const reference=await inventoryOf(page,'#main');
+    await restore();
+    await openProduction(page,`collection/${manifest.collections.Landscapes}`,theme);await page.mouse.move(0,0);
+    await expect(page.locator(CARDS)).toHaveCount(12);
+    await expect(page.locator('#workspace-main #page-title')).toHaveText('Landscapes');
+    await expect(page.locator('#workspace-breadcrumb')).toHaveText('Collection');
+    await expect(page).toHaveTitle('Collection · Prompt Manager');
+    await settle(page);
+    const actual=await inventoryOf(page,'#workspace-main');
+    expect.soft(actual.heading.actions.buttons.map(item=>item.text),'production heading actions').toEqual(['Edit','Import','New prompt']);
+    expect.soft(reference.heading.actions.buttons.map(item=>item.text),'mapped concept heading actions').toEqual(['Edit','Import','New prompt']);
+    expect.soft(compare(reference,actual,manifest,['heading','toolbar','filterRow']),'collection page').toEqual([]);
+    // The comparison notices a lost Edit.
+    await mutate(page,{remove:['#workspace-main .heading-actions > button:first-child']});
+    expect(compare(reference,await inventoryOf(page,'#workspace-main'),manifest,['heading']).length).toBeGreaterThan(0);
+});
+
 /** #332: an own category on one listed image, set through the real route (PUT /prompt_manager/images/category). */
 const IMAGE_CATEGORY='Portraits';
 async function setImageCategory(request,path,category) {

tests/browser/inspector-concept.spec.mjs

@@ -1,7 +1,7 @@
 import {writeFile} from 'node:fs/promises';
 import {test,expect} from './fixtures.mjs';
 import {captureCanonicalState} from './history-canonical-comparison.mjs';
-import {SIZES,THEMES,seedConcept,openConcept,openProduction,inspectorOf,diffInventory,mutate,countCopy,saveImageTags,withChipSource} from './concept-inventory.mjs';
+import {SIZES,THEMES,seedConcept,openConcept,openProduction,inspectorOf,diffInventory,mutate,countCopy,saveImageTags,withChipSource,withTagSections,PROMPT_TAGS} from './concept-inventory.mjs';
 
 // Task D4 of #249: the Images side panel (app.js imageDetails a400–414) and the viewer's full details (imageDetails(i,
 // true)) against the immutable concept, on the concept-shaped real fixture, at 1600x1100 and 390x844 in both themes.
@@ -51,10 +51,12 @@ function withCategoryScope(inventory) {
 }
 
 /**
- * Recorded inspector deviations only (spec §8, TESTING.md D3, D20–D22, D34, D38, D117 and D146); everything else must
+ * Recorded inspector deviations only (spec §8, TESTING.md D3, D20–D22, D34, D38, D117, D139, D146 and D154, whose labelled
+ * tag sections the concept page gets at runtime from `withTagSections`); everything else must
  * equal production. `reference.viewerPlain` is the concept's own full details
  * with its tag chips made plain (D20), D117's summary button added and, under a coarse pointer, D34's targets, all at
- * runtime: the browser lays out what follows them, as the Images spec's removed toolbar items do.
+ * runtime: the browser lays out what follows them, as the Images spec's removed toolbar items do. D139's
+ * "Collections…" is drawn after Prompt history in both the concept's panel footer and its full details.
  */
 export function mapInspectorConcept(reference,{viewport}) {
     const {panel,viewerPlain,fileControls}=reference;
@@ -107,18 +109,36 @@ async function openProductionPanel(page,theme) {
     await expect(page.locator(CARDS).first()).toBeVisible();
     await expect(panelOf(page).locator('.panel-heading h2')).toHaveText(TITLE);
     await expect(panelOf(page).locator('.detail-footer')).toBeVisible();
-    await expect(panelOf(page).locator('.detail-inner > .tags')).toHaveCount(2);
+    await expect(panelOf(page).locator('.detail-inner > .tags')).toHaveCount(1);
+    await expect(panelOf(page).locator('.detail-inner > .tag-section.linked-tags')).toBeVisible();
     await settle(page);
 }
 
 /**
- * The concept's tag chips in the open viewer as plain `span.chip`s (D20), D34's coarse targets and D117's "Copy
- * settings summary" after the six facts (LI-4), drawn as the concept's own plain button (its "Download sample JSON",
- * without its action); returns an undo.
+ * #331 D139: production's "Collections…" after Prompt history in `selector`'s buttons, drawn in the concept as its own
+ * Prompt history button (no action) with the concept's folder icon (its Images filter chip's); returns an undo.
  */
-async function plainViewerChips(page) {
+async function withCollectionsButton(page,selector) {
+    await page.evaluate(selector=>{
+        const history=[...document.querySelectorAll(`${selector} > button`)].find(button=>button.textContent.includes('Prompt history'));
+        const button=history.cloneNode(false);
+        for(const key of Object.keys(button.dataset))delete button.dataset[key];
+        button.append(document.querySelector('.filter-row .chip > svg.icon').cloneNode(true),' Collections…');
+        history.after(button);(window.__collectionsButtons??=[]).push(button);
+    },selector);
+    await settle(page);
+    return async()=>{await page.evaluate(()=>{for(const button of window.__collectionsButtons??[])button.remove();delete window.__collectionsButtons;});};
+}
+
+/**
+ * The concept's tag chips in the open viewer in D154's labelled sections (`sections`) as plain `span.chip`s (D20), D34's
+ * coarse targets and D117's "Copy settings summary" after the six facts (LI-4), drawn as the concept's own plain button
+ * (its "Download sample JSON", without its action); returns an undo.
+ */
+async function plainViewerChips(page,sections) {
     const restoreTargets=await mutate(page,{css:D34_COARSE_TARGETS});
     const restoreSource=await withChipSource(page,'#viewer-details',ORIGIN);
+    const restoreSections=await withTagSections(page,'#viewer-details',sections);
     await page.evaluate(()=>{
         window.__plainChips=[...document.querySelectorAll('#viewer-details .tags > button.chip')].map(button=>{
             const span=document.createElement('span');span.className='chip';span.textContent=button.textContent;
@@ -129,11 +149,12 @@ async function plainViewerChips(page) {
         summary.textContent='Copy settings summary';summary.classList.add('settings-summary');
         document.querySelector('#viewer-details > div > dl.meta-grid').after(summary);window.__summaryButton=summary;
     });
-    await settle(page);
+    const restoreCollections=await withCollectionsButton(page,'#viewer-details');
     return async()=>{
+        await restoreCollections();
         await page.evaluate(()=>{for(const [span,button] of window.__plainChips)span.replaceWith(button);delete window.__plainChips;
             window.__summaryButton.remove();delete window.__summaryButton;});
-        await restoreSource();await restoreTargets();
+        await restoreSections();await restoreSource();await restoreTargets();
     };
 }
 
@@ -151,7 +172,7 @@ async function fileControlsOf(page) {
  * The first card's viewer: its full details and open "File & provenance" measured (in the concept also with plain
  * chips), then closed again. `mutation` (production) is applied once the viewer is open and undone before it closes.
  */
-async function viewerOf(page,concept,mutation=null) {
+async function viewerOf(page,concept,mutation=null,sections=PROMPT_TAGS) {
     await page.locator(CARDS).first().locator('.image-open').click();
     await expect(page.locator('#viewer-dialog[open] #viewer-details .panel-heading h2')).toHaveText(TITLE);
     await settle(page);
@@ -160,7 +181,7 @@ async function viewerOf(page,concept,mutation=null) {
     const {viewerDetails}=await inspectorOf(page);
     const fileControls=await fileControlsOf(page);
     let viewerPlain=null;
-    if(concept) {const restore=await plainViewerChips(page);viewerPlain=(await inspectorOf(page)).viewerDetails;await restore();}
+    if(concept) {const restore=await plainViewerChips(page,sections);viewerPlain=(await inspectorOf(page)).viewerDetails;await restore();}
     await restoreMutation?.();
     await page.keyboard.press('Escape');
     await expect(page.locator('#viewer-dialog[open]')).toHaveCount(0);
@@ -168,13 +189,18 @@ async function viewerOf(page,concept,mutation=null) {
     await page.evaluate(()=>new Promise(resolve=>setTimeout(resolve,0)));await settle(page);
     return {viewerDetails,viewerPlain,fileControls};
 }
-async function inventory(page,concept=false) {
-    // D146: the concept panel's category chip with its visible origin, laid out by the concept page itself.
+async function inventory(page,concept=false,sections=PROMPT_TAGS) {
+    // D146: the concept panel's category chip with its visible origin; D139: Collections… after Prompt history; D154's
+    // labelled tag sections (by default the fixture's D39 positive tags under "Prompt tags"), laid out by the concept page.
     const restoreSource=concept?await withChipSource(page,'#image-inspector',ORIGIN):null;
+    const restoreCollections=concept?await withCollectionsButton(page,'#image-inspector > .detail-footer'):null;
+    const restoreSections=concept?await withTagSections(page,'#image-inspector',sections):null;
     if(concept)await settle(page);
     const {panel}=await inspectorOf(page);
+    await restoreSections?.();
+    await restoreCollections?.();
     await restoreSource?.();
-    const {viewerDetails,viewerPlain,fileControls}=await viewerOf(page,concept);
+    const {viewerDetails,viewerPlain,fileControls}=await viewerOf(page,concept,null,sections);
     return concept?{panel,viewerDetails,viewerPlain,fileControls}:{panel,viewerDetails,fileControls};
 }
 
@@ -238,7 +264,8 @@ async function productionSendOf(page,context) {
 /** A tag chip: the collection it leaves (the count line), then the chip pressed again for the same result. */
 async function tagOf(page) {
     const before=await page.locator('#result-count').textContent();
-    await panelOf(page).locator('.detail-inner > .tags').last().getByRole('button',{name:'cinematic',exact:true}).click();
+    // The concept's bare row, or production's linked-tags section (D154): the last chip row either way.
+    await panelOf(page).locator('.detail-inner .tags').last().getByRole('button',{name:'cinematic',exact:true}).click();
     // Production's caption reads "Loading images…" until the tag's listing answers (#391); the concept's search is instant.
     await expect.poll(async()=>{const text=await page.locator('#result-count').textContent();return text!==before&&text!=='Loading images…';}).toBe(true);
     await settle(page);
@@ -303,6 +330,9 @@ for(const [width,height] of SIZES)for(const theme of THEMES)test(`inspector pane
     expect.soft(panel.badges.chips.map(chip=>chip.text),'concept badges').toEqual(['Execution recorded','Landscapes · prompt']);
     expect.soft(reference.viewerDetails.structure.inner.slice(-2),'concept full details').toEqual(['details','details']);
 
+    // D154: both documents measured the labelled section, so the comparison covers it (never two absent sections).
+    for(const [name,inventory] of [['concept panel',reference.panel],['concept viewer',reference.viewerPlain],['panel',actual.panel],['viewer',actual.viewerDetails]])
+        expect.soft(inventory.tagSections.map(section=>section.label.text),`${name} tag sections`).toEqual(['Prompt tags · 3']);
     expect.soft(diff,'differences after the recorded deviations').toEqual([]);
     // Copy: the concept's toast "Prompt copied." (or its dialog), in both documents.
     expect.soft(production.copy,'copy').toEqual(concept.copy);
@@ -332,15 +362,18 @@ test('a recipe-linked image without recipe metadata shows its positive prompt me
     await openProductionPanel(page,'dark');
     await page.locator(CARDS).nth(1).locator('[aria-label^="Inspect"]').click();
     await expect(panelOf(page).locator('.panel-heading h2')).toHaveText('Quiet geometry');
-    const rows=panelOf(page).locator('.detail-inner > .tags');
-    await expect(rows).toHaveCount(2);
-    await expect(rows.first().locator('.chip')).toHaveText(['Execution recorded','Positive prompt category: Architecture · prompt']);
-    await expect(rows.first().locator('.chip .ca-visually-hidden')).toHaveText('Positive prompt category: ');
+    const badges=panelOf(page).locator('.detail-inner > .tags');
+    await expect(badges).toHaveCount(1);
+    await expect(badges.locator('.chip')).toHaveText(['Execution recorded','Positive prompt category: Architecture · prompt']);
+    await expect(badges.locator('.chip .ca-visually-hidden')).toHaveText('Positive prompt category: ');
     // D146 (#332): the visible origin names where the value came from.
-    await expect(rows.first().locator('.chip .chip-source')).toHaveText(' · prompt');
-    await expect(rows.nth(1).locator('button.chip')).toHaveText(['minimal','warm_light','editorial']);
-    await expect(rows.nth(1).locator('span.chip')).toHaveCount(0);
-    await expect(panelOf(page).locator('.detail-footer button')).toHaveText(['Send to ComfyUI','Prompt history']);
+    await expect(badges.locator('.chip .chip-source')).toHaveText(' · prompt');
+    // D154: the positive's tags, so the section is "Prompt tags", not "Recipe tags".
+    const tags=panelOf(page).getByRole('region',{name:'Prompt tags',exact:true});
+    await expect(tags.locator('.role-label')).toHaveText('Prompt tags · 3');
+    await expect(tags.locator('.tags > button.chip')).toHaveText(['minimal','warm_light','editorial']);
+    await expect(tags.locator('span.chip')).toHaveCount(0);
+    await expect(panelOf(page).locator('.detail-footer button')).toHaveText(['Send to ComfyUI','Prompt history','Collections…']);
 });
 
 // #391: a tag chip replaces the collection; until its listing answers the caption says it is loading, never that the
@@ -353,7 +386,7 @@ test('a tag chip collection reads as loading, not empty, until its listing answe
     await page.route(url=>url.pathname==='/prompt_manager/images/output'&&url.searchParams.get('tags')==='cinematic',
         async route=>{arrived();await held;await route.continue();},{times:1});
     const count=page.locator('#result-count');
-    await panelOf(page).locator('.detail-inner > .tags').last().getByRole('button',{name:'cinematic',exact:true}).click();
+    await panelOf(page).locator('.tag-section.linked-tags').getByRole('button',{name:'cinematic',exact:true}).click();
     await reached;
     await expect(count).toHaveText('Loading images…');
     await expect(page.locator(CARDS)).toHaveCount(0);
@@ -363,8 +396,9 @@ test('a tag chip collection reads as loading, not empty, until its listing answe
 });
 
 // Task C3 (D88, #249 PR C): the concept's chip row is the image's own tags (`i.tags`, app.js a402). With those names
-// saved as the file's current image tags, production draws the row from them (Tags filter buttons in the panel, plain
-// in the viewer, the linked positive's equal tags not repeated) and still equals the concept, in both sizes.
+// saved as the file's current image tags, production draws them in D154's "Image tags" section (Tags filter buttons in
+// the panel, plain in the viewer) and the linked positive's equal tags in its "Prompt tags" section after it (#449),
+// and equals the concept with its row in those two sections, in both sizes.
 for(const [width,height] of SIZES)test(`image tags saved on the concept image keep the inspector equal to the concept ${width}x${height} (D88)`,async({page,request},testInfo)=>{
     test.setTimeout(120_000);
     const viewport={width,height};await page.setViewportSize(viewport);
@@ -372,13 +406,14 @@ for(const [width,height] of SIZES)test(`image tags saved on the concept image ke
     const names=await panelOf(page).locator('.detail-inner > .tags').last().locator('button.chip').allTextContents();
     expect(names).toEqual(['cinematic','blue_hour','architecture']);
     await captureCanonicalState(page,testInfo,`concept-image-tags-${width}`,'#image-inspector');
-    const reference=await inventory(page,true);
+    const reference=await inventory(page,true,[['image-tags','Image tags','Image tags'],...PROMPT_TAGS]);
     const {path}=testInfo.manifest.images.find(item=>item.filename==='alpine_00001.png');
     expect(await saveImageTags(request,path,names)).toEqual(names);
     const read=page.waitForResponse(item=>new URL(item.url()).pathname===`/prompt_manager/images/inspect/${path}`);
     await openProductionPanel(page,'dark');
     expect((await (await read).json()).image_tags).toEqual({state:'current',tags:names});
-    await expect(panelOf(page).locator('.detail-inner > .tags').last().locator('.chip')).toHaveText(names);
+    await expect(panelOf(page).locator('.tag-section.image-tags .chip')).toHaveText(names);
+    await expect(panelOf(page).locator('.tag-section.linked-tags .chip')).toHaveText(names);
     await captureCanonicalState(page,testInfo,`production-image-tags-${width}`,'#image-inspector');
     assertInspectorMatch(reference,await inventory(page),viewport);
 });
@@ -403,6 +438,17 @@ test('inspector comparison detects mutations',async({page})=>{
     await restoreCss();await settle(page);
     assertInspectorMatch(reference,await measure(),viewport);
 
+    // D154: the tag section's muted label and its row's spacing are part of the mapped concept.
+    for(const [css,path] of [['#image-inspector .tag-section .role-label { color: rgb(255, 0, 0) }',/panel\.tagSections\.0\.label\.color/],
+        ['#image-inspector .tag-section > .tags { margin-top: 12px }',/panel\.tagSections\.0\.row\.marginTop/],
+        ['#image-inspector .tag-section { border-top-width: 0 }',/panel\.tagSections\.0\.borderTopWidth/]]) {
+        const restore=await mutate(page,{css});
+        const mutated=await measure();
+        expect(()=>assertInspectorMatch(reference,mutated,viewport)).toThrow(path);
+        await restore();await settle(page);
+    }
+    assertInspectorMatch(reference,await measure(),viewport);
+
     const panel=(await inspectorOf(page)).panel;
     const viewerMutated=async mutation=>{const {viewerDetails,fileControls}=await viewerOf(page,false,mutation);return {panel,viewerDetails,fileControls};};
     const withoutActions=await viewerMutated({remove:['#viewer-details > button']});
@@ -414,6 +460,12 @@ test('inspector comparison detects mutations',async({page})=>{
     expect(()=>assertInspectorMatch(reference,withoutSummary,viewport)).toThrow(/viewerDetails\.structure\.inner/);
     assertInspectorMatch(reference,await viewerMutated(),viewport);
 
+    // D139: Collections… is part of the mapped concept, so losing it is a difference.
+    const restoreButton=await mutate(page,{remove:['#image-inspector > .detail-footer > button:last-child']});
+    const withoutCollections=await measure();
+    expect(()=>assertInspectorMatch(reference,withoutCollections,viewport)).toThrow(/panel\.footer/);
+    await restoreButton();await settle(page);
+
     // D146: the visible origin is part of the mapped concept, so losing it is a difference.
     const restoreSource=await mutate(page,{remove:['#image-inspector .chip-source']});
     const sourceMutated=await measure();

tests/browser/library-concept.spec.mjs

@@ -21,8 +21,8 @@ export function mapLibraryConcept(reference) {
     // Save view is the concept's own (saved views, #249 D16 retired).
     const toolbar={...library.toolbar};
     toolbar.items=toolbar.items.map(item=>{
-        // D14: the Library appends its Oldest first after the concept's three sorts.
-        if(item.id==='sort')return {...item,options:[...item.options,'Oldest first']};
+        // D14: the Library appends its Oldest first after the concept's sorts; D153 (#447) drops "Highest rated".
+        if(item.id==='sort')return {...item,options:[...item.options.filter(option=>option!=='Highest rated'),'Oldest first']};
         // D36: the category list is the Library's real list, sorted alphabetically; the concept lists prompt order.
         if(item.id==='category-filter')return {...item,options:[item.options[0],...item.options.slice(1).sort(nocase)]};
         return item;
@@ -30,17 +30,20 @@ export function mapLibraryConcept(reference) {
     // A shorter toolbar moves what follows it up by as much.
     const rise=library.toolbar.height-toolbar.height;
     const cards=library.cards.map((card,index)=>({...card,
+        // D148 (#437): each card leads with the concept's Images `.image-overlays` selection checkbox, placed absolutely
+        // over the art, so nothing else moves; it is compared with the concept's own overlay below.
+        children:['image-overlays',...card.children],
         // D23: "− N negative variants" uses the singular for one; that row is the concept's own, measured so (`oracle`).
         meta:card.meta.items[0]?.text===PLURAL_ONE?reference.oracle.meta[index]:card.meta,
         // D23: a long title wraps inside the title row, so the title button may shrink below its text (`min-width: 0`).
         title:{...card.title,items:card.title.items.map(item=>item.tag==='button'?{...item,minWidth:'0px'}:item)},
         // D43: tag chips come in the store's alphabetical order; the row flows again from the concept's own chips.
         tags:sortedTags(card.tags),
-        // D23: the stars are aria-hidden, with a visually hidden "Rated N out of 5" spoken in their place. The fixture's
-        // run_count (4) and last_used_at (its own "now", where the page's clock stands) give the concept's own "4 runs"
-        // and "Used today". The two-line clamp is
+        // D153 (#447): no rating stars; the concept page itself is measured without them (`dropConceptRatings`). The
+        // fixture's run_count (4) and last_used_at (its own "now", where the page's clock stands) give the concept's own
+        // "4 runs" and "Used today". The two-line clamp is
         // applied below through `clampedCards`, measured on the concept with `CLAMP_CSS` (its samples wrap at 390px).
-        footer:{...card.footer,facts:{...card.footer.facts,spoken:card.footer.facts.spoken.replace(/^(★+)/,stars=>`Rated ${stars.length} out of 5`)}}}));
+        footer:card.footer}));
     return {...reference,library:{...library,toolbar,
         // D35: the caption uses the singular for one.
         caption:{...library.caption,top:library.caption.top-rise,spans:library.caption.spans.map(span=>({...span,text:countCopy(span.text)}))},
@@ -136,9 +139,20 @@ async function sortConceptTags(page) {
     return ()=>page.evaluate(()=>{for(const [row,chips] of window.__conceptTagOrder)row.append(...chips);delete window.__conceptTagOrder;});
 }
 
+/**
+ * D153 (#447): production draws no 1–5 rating. The concept's card footers read "★★★★★ · Used today"; the stars and
+ * their separator are taken out of the concept page itself, so the footer facts are measured as production draws them.
+ * Returns the concept's original footer texts.
+ */
+async function dropConceptRatings(page) {
+    return page.evaluate(()=>[...document.querySelectorAll('.prompt-card .card-footer > span')].map(span=>{
+        const original=span.textContent;span.textContent=original.replace(/^★+ · /,'');return original;}));
+}
 async function openConceptLibrary(page,theme,motion) {
     await openConcept(page,'library',theme,motion);
     await expect(page.locator(CARDS)).toHaveCount(CONCEPT_PROMPTS);
+    const footers=await dropConceptRatings(page);
+    expect(footers,'the concept rates every card (D153 removes it)').toEqual(Array(CONCEPT_PROMPTS).fill().map((_,index)=>expect.stringMatching(/^★+ · \S/)));
     await page.mouse.move(0,0);await decoded(page);await settle(page);
 }
 async function openProductionLibrary(page,theme,motion) {
@@ -292,7 +306,7 @@ for(const [width,height] of SIZES)for(const theme of THEMES)test(`library geomet
             expect.soft(card.meta,`${motion} concept meta`).toMatchObject({fontSize:'10px',columnGap:'9px'});
             expect.soft(card.meta.items.map(item=>item.text),`${motion} concept meta text`).toEqual(['− 3 negative variants','4 runs','4 images']);
             expect.soft(card.footer,`${motion} concept footer`).toMatchObject({paddingTop:'11px',paddingRight:'18px',paddingBottom:'11px',paddingLeft:'18px',borderTopWidth:'1px',borderTopStyle:'solid'});
-            expect.soft(card.footer.facts.text,`${motion} concept footer facts`).toBe('★★★★★ · Used today');
+            expect.soft(card.footer.facts.text,`${motion} concept footer facts (D153: without the stars)`).toBe('Used today');
             expect.soft(diffInventory(stripExact(cards.map(item=>sortedTags(item.tags))),stripExact(sorted)),`${motion} tags in alphabetical order`).toEqual([]);
             expect.soft(diffInventory(stripExact(clampedCards(reference.library)),stripExact({grid:clamped.grid,cards:clamped.cards})),`${motion} clamped cards`).toEqual([]);
             expect.soft(diff,`${motion} differences after the recorded deviations`).toEqual([]);
@@ -311,6 +325,56 @@ for(const [width,height] of SIZES)for(const theme of THEMES)test(`library geomet
     } finally {await host.evaluate(()=>window.conceptHost.dispose());await host.close();}
 });
 
+/**
+ * D148 (#437): the selection checkbox overlay of `card` and the `.bulkbar` row inside `main`, with one item checked,
+ * measured alike in the concept's Images page and production's Library: the overlay's inset in its card, the checkbox,
+ * and the bar's box, tokens and the actions both bars have (Add tags, Set category, Clear; #442).
+ */
+const measureSelection=(page,{card,main})=>page.evaluate(({card,main})=>{
+    const css=(node,...keys)=>{const style=getComputedStyle(node);return Object.fromEntries(keys.map(key=>[key,style[key]]));};
+    const rect=node=>node.getBoundingClientRect(),px=value=>Math.round(value*100)/100;
+    const font=node=>css(node,'fontFamily','fontSize','fontWeight','lineHeight','color');
+    const edges=node=>css(node,'paddingTop','paddingRight','paddingBottom','paddingLeft','marginBottom','borderTopWidth','borderTopStyle',
+        'borderTopColor','borderTopLeftRadius','backgroundColor');
+    const host=document.querySelector(card),overlay=host.querySelector('.image-overlays'),label=overlay.querySelector(':scope > label');
+    const input=label.querySelector('input'),root=document.querySelector(main),bar=root.querySelector('.bulkbar');
+    const shared=['Add tags','Set category','Clear'];
+    return {
+        overlay:{left:px(rect(overlay).left-rect(host).left),top:px(rect(overlay).top-rect(host).top),right:px(rect(host).right-rect(overlay).right),
+            ...css(overlay,'position','display','justifyContent','pointerEvents')},
+        label:css(label,'pointerEvents'),checkbox:{width:px(rect(input).width),height:px(rect(input).height),checked:input.checked,...css(input,'accentColor')},
+        bar:{left:px(rect(bar).left-rect(root).left),...css(bar,'display','alignItems','columnGap','rowGap','flexWrap'),...edges(bar),...font(bar),
+            count:{text:bar.querySelector(':scope > strong').textContent,...font(bar.querySelector(':scope > strong'))},
+            buttons:[...bar.querySelectorAll(':scope > button')].filter(button=>shared.includes(button.textContent.trim()))
+                .map(button=>({text:button.textContent.trim(),height:px(rect(button).height),...font(button),...edges(button)}))},
+    };
+},{card,main});
+async function conceptSelection(page,theme) {
+    await openConcept(page,'images',theme);
+    await page.locator('#image-grid .media-card').first().locator('input[data-select]').check();
+    await expect(page.locator('#bulkbar .bulkbar')).toBeVisible();await page.mouse.move(0,0);await settle(page);
+    return measureSelection(page,{card:'#image-grid .media-card',main:'#main'});
+}
+async function productionSelection(page,theme) {
+    await openProductionLibrary(page,theme,'no-preference');
+    await page.locator(CARDS).first().locator('.image-overlays input').check();
+    await expect(page.locator('#library-bulkbar .bulkbar')).toBeVisible();await page.mouse.move(0,0);await settle(page);
+    return measureSelection(page,{card:CARDS,main:'#workspace-main'});
+}
+for(const [width,height] of SIZES)for(const theme of THEMES)test(`the Library selection checkbox and bar match the concept's Images ones ${width}x${height} ${theme}`,async({page},testInfo)=>{
+    await page.setViewportSize({width,height});
+    const reference=await conceptSelection(page,theme);
+    await captureCanonicalState(page,testInfo,`concept-images-selection-${width}-${theme}`);
+    const actual=await productionSelection(page,theme);
+    await captureCanonicalState(page,testInfo,`production-library-selection-${width}-${theme}`);
+    await testInfo.attach('library-selection',{body:JSON.stringify({reference,actual},null,2),contentType:'application/json'});
+    expect(reference.bar.buttons.map(button=>button.text),'the concept bar has every Library action').toEqual(['Add tags','Set category','Clear']);
+    expect(actual).toEqual(reference);
+    // The comparison notices a moved checkbox (spec §7 mutation check).
+    await mutate(page,{css:'.prompt-card .image-overlays { inset: 20px 9px auto }'});
+    expect((await measureSelection(page,{card:CARDS,main:'#workspace-main'})).overlay.top).not.toBe(reference.overlay.top);
+});
+
 // The grid and cards at the concept's three-column breakpoint (style.css 1274–1276).
 for(const theme of THEMES)test(`library grid and cards match the concept 1800x1100 ${theme}`,async({page},testInfo)=>{
     await page.setViewportSize({width:1800,height:1100});
@@ -341,7 +405,7 @@ test('a run that crosses midnight still reads the concept fixture\'s day',async(
     const reference=await conceptInventory(page);
     await openProductionLibrary(page,'dark','no-preference');
     const actual=await inventoryOf(page,'#workspace-main');
-    expect(actual.library.cards.map(card=>card.footer.facts.text.replace(/^★+ · /,''))).toEqual(Array(CONCEPT_PROMPTS).fill('Used today'));
+    expect(actual.library.cards.map(card=>card.footer.facts.text)).toEqual(Array(CONCEPT_PROMPTS).fill('Used today'));
     expect(compare(reference,actual,{sections:['library'],library:['grid','cards','trailing']})).toEqual([]);
 });
 

tests/browser/library-editor-canonical-comparison.mjs

@@ -25,9 +25,10 @@ export function inventoryOf(dialog) {
         demo:Boolean(dialog.querySelector('.demo-note')),
     };
 }
-// Documented differences only: prototype Save label, prototype demo note, and data-dependent values.
+// Documented differences only: prototype Save label, prototype demo note, data-dependent values, and D153 (#447): the
+// concept's "Rating" select is not drawn.
 export function mapConcept(reference) {
-    return {...reference,demo:false,actions:{...reference.actions,buttons:reference.actions.buttons.map(item=>item.text==='Save in prototype'?{...item,text:'Save'}:item)}};
+    return {...reference,demo:false,fields:reference.fields.filter(item=>item.label!=='Rating'),actions:{...reference.actions,buttons:reference.actions.buttons.map(item=>item.text==='Save in prototype'?{...item,text:'Save'}:item)}};
 }
 
 // Flatten an inventory to path->value so a comparison reports exactly which leaves differ.

tests/browser/prompt-inspector.spec.mjs

@@ -1,5 +1,6 @@
 import {startCoverage,saveCoverage} from '../../tools/vendor/browser-coverage.mjs';
 import {test,expect} from './fixtures.mjs';
+import { startInGrid } from './concept-inventory.mjs';
 
 const workspace='/prompt_manager/workspace#/images';
 const details=page=>page.getByRole('complementary',{name:'Selected image details',exact:true});
@@ -252,6 +253,8 @@ test('responsive gallery defers observed size writes and still fits the new view
             });}
         };
     });
+    // The Grid's observed sizes (D157: the default layout is Masonry).
+    await startInGrid(page);
     await page.setViewportSize({width:1280,height:900});await page.goto(workspace);
     await expect(page.locator('.media-card')).toHaveCount(4);
     await page.setViewportSize({width:390,height:844});

tests/browser/system-concept.spec.mjs

@@ -162,7 +162,6 @@ async function serverJson(request, url) {
 }
 
 const number = value => value.toLocaleString('en-US');
-const ratingText = value => value === null ? 'Not rated' : `${number(value)} / 5`;
 
 /** The concept's shell and Overview with the recorded deviations applied, given the server's real overview. */
 function mapOverviewConcept(reference, overview) {
@@ -178,8 +177,8 @@ function mapOverviewConcept(reference, overview) {
         'One display configuration': row => ({ ...row, description: 'Grid and density settings are shared across views.' }),
         // D93: the real sources, not samples.
         '2 sample output sources': row => ({ ...row, title: `${number(overview.sources.effective)} output sources` }),
-        // D96: the dialog shows runs and ratings by category and library totals, not prompt, tag or date usage.
-        Statistics: row => ({ ...row, description: 'Runs and ratings by category, with library totals.' }),
+        // D96: the dialog shows runs by category and library totals, not prompt, tag or date usage (D153: no ratings).
+        Statistics: row => ({ ...row, description: 'Runs by category, with library totals.' }),
     };
     const surfaces = reference.columns.surfaces.map(surface => ({ ...surface,
         rows: surface.rows.map(row => (rows[row.title] ?? (value => value))(row)) }));
@@ -288,7 +287,10 @@ test('system comparison detects mutations', async ({ page, request }) => {
     expect(diffInventory(expected, await measure())).toEqual([]);
 });
 
-/** D96: the concept's table, its rows the real ones; then "Library totals" in the same table markup. */
+/**
+ * D96: the concept's table, its rows the real ones; then "Library totals" in the same table markup. D153 (#447): the
+ * concept's "Avg. rating" column and a "Rated images" total are not drawn.
+ */
 function mapStatisticsConcept(reference, statistics) {
     const [runs] = reference.tables;
     return {
@@ -296,10 +298,11 @@ function mapStatisticsConcept(reference, statistics) {
         // No sample note (prototype copy); a second table follows the first.
         body: ['table', 'table'],
         tables: [
-            { ...runs, rows: statistics.by_category.map(row => [row.category ?? 'Uncategorized', number(row.runs), ratingText(row.average_rating)]) },
+            { ...runs, head: runs.head.filter(([, , text]) => text !== 'Avg. rating'),
+                rows: statistics.by_category.map(row => [row.category ?? 'Uncategorized', number(row.runs)]) },
             { caption: 'Library totals', head: [['th', 'col', 'Total'], ['th', 'col', 'Count']],
                 rows: [['Prompts', number(statistics.totals.total_prompts)], ['Tags', number(statistics.totals.total_tags)],
-                    ['Images', number(statistics.totals.total_images)], ['Rated images', number(statistics.rated_images)]] },
+                    ['Images', number(statistics.totals.total_images)]] },
         ],
     };
 }

tests/browser/unseen-restore.mjs

@@ -52,3 +52,43 @@ export async function unseenRestore(page, restoring) {
         open();
     }
 }
+
+// A restore the open page does see run and finish (#415), driven by the test instead of the 15 s poll's timing. The
+// job is shaped as the server's (py/api/backup_restore.py: kind `restore`, one `errors` group, not cancellable).
+const RESTORE_JOB_ID = '415415415415415415415415415415ab';
+const restoreJob = state => ({ id: RESTORE_JOB_ID, kind: 'restore', title: 'Database restore', state, started_at: 1,
+    finished_at: state === 'running' ? null : 2, progress: 0, status_line: state === 'running' ? '' : 'Restored the test snapshot.',
+    processed: 0, total: 0, counters: { errors: 0 }, scope: {}, listed: { errors: 0 }, log_length: 0, cancellable: false });
+/** The Activity pill's link while the page follows that restore. */
+export const SEEN_RESTORE_PILL = `#/activity/${RESTORE_JOB_ID}`;
+/** The toast the page shows when it sees that restore finish (activity-kinds.mjs `restore.finishToast`). */
+export const SEEN_RESTORE_TOAST = 'Database restore finished · Restored the test snapshot.';
+
+/**
+ * Call before the page's first navigation, after `armUnseenRestores(page)` (a later route wins). The page's first
+ * job-list read lists a restore job running, so its Activity store follows it; its job stream is refused, so the store
+ * reads the list again, and every later read is held until `finish()`, which answers them with the job finished and
+ * resolves once one of those answers reached the page. The real restore the test runs meanwhile goes through the API.
+ */
+export async function followRestore(page) {
+    let release;
+    const released = new Promise(resolve => { release = resolve; });
+    let first = true;
+    await page.route(`${JOBS}/*/events`, route => route.abort().catch(() => {}));
+    await page.route(JOBS, async route => {
+        if (first) {
+            first = false;
+            return route.fulfill({ json: { success: true, jobs: [restoreJob('running')] } }).catch(() => {});
+        }
+        await released;
+        // A read the page has aborted since (a newer read supersedes it) can no longer be answered.
+        return route.fulfill({ json: { success: true, jobs: [restoreJob('finished')] } }).catch(() => {});
+    });
+    return {
+        async finish() {
+            const answered = page.waitForResponse(response => new URL(response.url()).pathname === '/prompt_manager/jobs');
+            release();
+            await answered;
+        },
+    };
+}

tests/browser/workspace-collections.spec.mjs

@@ -491,3 +491,366 @@ test('a page that says a collection or view could not be opened takes focus afte
     await expect(description(page)).toHaveText('Saved views could not be loaded. Your saved data has not changed.');
     await expect(heading(page)).toBeFocused();
 });
+
+// #331 D138 (design §7): the selection bar's Add to collection and Remove from collection.
+const bar = page => page.locator('#bulkbar > .bulkbar');
+const barButton = (page, name) => bar(page).getByRole('button', { name, exact: true });
+const addDialog = page => page.getByRole('dialog', { name: 'Add selected images to collection', exact: true });
+const addField = page => addDialog(page).getByRole('combobox', { name: 'Collection name', exact: true });
+const check = (page, file) => page.getByRole('checkbox', { name: `Select ${file}`, exact: true }).check();
+const card = (page, path) => page.locator(`article[data-image-key$=":${path}"]`);
+
+/** Add the checked images under `name` through the bar's dialog. */
+async function addChecked(page, name) {
+    await barButton(page, 'Add to collection').click();
+    await expect(addDialog(page)).toBeVisible();
+    await addField(page).fill(name);
+    await addDialog(page).getByRole('button', { name: 'Add', exact: true }).click();
+    await expect(addDialog(page)).toHaveCount(0);
+}
+
+test('Add to collection creates a collection and adds to it ignoring case (CUJ 1–3)', async ({ page, request }) => {
+    await seed(request);
+    await page.goto(`${workspace}#/images`);
+    await expect(imageCards(page)).toHaveCount(21);
+    for (const file of ['loose.png', 'landscape_01.png', 'experiment_01.png']) await check(page, file);
+    await expect(bar(page).locator('button')).toHaveText(['Add tags', 'Auto-tag selected', 'Set category', 'Add to collection',
+        'Export metadata', 'Clear']);
+    await barButton(page, 'Add to collection').click();
+    // The concept's simpleInput with Add (D138) and the existing names as suggestions.
+    await expect(addDialog(page).locator('label')).toHaveText('Collection name');
+    await expect(addField(page)).toHaveValue('');
+    await expect(addField(page)).toHaveAttribute('maxlength', '100');
+    await expect(addDialog(page).locator('datalist option')).toHaveCount(2);
+    expect(await addDialog(page).locator('datalist option').evaluateAll(options => options.map(option => option.value)))
+        .toEqual(['Experiments', 'Landscapes']);
+    await expect(addDialog(page).locator('.dialog-actions button')).toHaveText(['Cancel', 'Add']);
+    // An empty name keeps the dialog on its field.
+    await addDialog(page).getByRole('button', { name: 'Add', exact: true }).click();
+    await expect(addField(page)).toBeFocused();
+    await addField(page).fill('  LoRa   testing ');
+    await addDialog(page).getByRole('button', { name: 'Add', exact: true }).click();
+    await expect(toast(page)).toHaveText('Added 3 images to “LoRa testing”.');
+    await expect(bar(page).locator('strong')).toHaveText('3 selected');
+    await expect(row(page, 'LoRa testing').locator('.nav-count')).toHaveText('3');
+    await expect(rows(page)).toHaveCount(3);
+
+    // CUJ 2: one of two already in it.
+    await bar(page).getByRole('button', { name: 'Clear', exact: true }).click();
+    await check(page, 'loose.png');
+    await check(page, 'landscape_02.png');
+    await addChecked(page, 'LoRa testing');
+    await expect(toast(page)).toHaveText('Added 1 image to “LoRa testing”. 1 was already in it.');
+    await expect(row(page, 'LoRa testing').locator('.nav-count')).toHaveText('4');
+
+    // CUJ 3: a case variant joins the same collection.
+    await bar(page).getByRole('button', { name: 'Clear', exact: true }).click();
+    await check(page, 'landscape_03.png');
+    await addChecked(page, 'lora TESTING');
+    await expect(toast(page)).toHaveText('Added 1 image to “LoRa testing”.');
+    await expect(rows(page)).toHaveCount(3);
+    await expect(row(page, 'LoRa testing').locator('.nav-count')).toHaveText('5');
+    const created = (await listed(request)).find(item => item.name === 'LoRa testing');
+    expect(created).toMatchObject({ count: 5, missing: 0 });
+
+    // Everything already in it.
+    await addChecked(page, 'LoRa testing');
+    await expect(toast(page)).toHaveText('All selected images were already in “LoRa testing”.');
+});
+
+test('Remove from collection takes the checked images off the open collection (CUJ 5)', async ({ page, request }) => {
+    const manifest = await seed(request);
+    const { id } = manifest.collections.Landscapes;
+    await page.goto(`${workspace}#/collection/${id}`);
+    await expect(imageCards(page)).toHaveCount(12);
+    await check(page, 'landscape_01.png');
+    await check(page, 'landscape_02.png');
+    await expect(bar(page).locator('button')).toHaveText(['Add tags', 'Auto-tag selected', 'Set category', 'Add to collection',
+        'Remove from collection', 'Export metadata', 'Clear']);
+    await barButton(page, 'Remove from collection').click();
+    await expect(toast(page)).toHaveText('Removed 2 images from “Landscapes”.');
+    await expect(imageCards(page)).toHaveCount(10);
+    await expect(card(page, 'collections/landscape_01.png')).toHaveCount(0);
+    // The removed files left the listing, so the bar has nothing checked.
+    await expect(bar(page)).toHaveCount(0);
+    await expect(row(page, 'Landscapes').locator('.nav-count')).toHaveText('10');
+    // A checked file still listed stays checked across the reload.
+    await check(page, 'landscape_03.png');
+    await check(page, 'landscape_04.png');
+    await page.route(`**${API}/items`, async route => {
+        if (route.request().method() !== 'DELETE') return route.fallback();
+        // Remove only landscape_04; landscape_03 stays listed and checked.
+        const body = route.request().postDataJSON();
+        await route.continue({ postData: JSON.stringify({ ...body, items: body.items.filter(item => item.path.endsWith('_04.png')) }) });
+    });
+    await barButton(page, 'Remove from collection').click();
+    await expect(toast(page)).toHaveText('Removed 1 image from “Landscapes”.');
+    await expect(imageCards(page)).toHaveCount(9);
+    await expect(bar(page).locator('strong')).toHaveText('1 selected');
+    await expect(page.getByRole('checkbox', { name: 'Select landscape_03.png', exact: true })).toBeChecked();
+    // No file changed: the plain Images page still lists all of them.
+    await page.goto(`${workspace}#/images`);
+    await expect(imageCards(page)).toHaveCount(21);
+});
+
+test('Images without a collection filter has no Remove from collection', async ({ page, request }) => {
+    const manifest = await seed(request);
+    await page.goto(`${workspace}#/images`);
+    await check(page, 'loose.png');
+    await expect(barButton(page, 'Add to collection')).toBeVisible();
+    await expect(barButton(page, 'Remove from collection')).toHaveCount(0);
+    // A saved view on a collection has its filter active: Remove shows (design §5).
+    await page.goto(`${workspace}#/saved/${manifest.saved_view.id}`);
+    await expect(imageCards(page)).toHaveCount(8);
+    await check(page, 'experiment_01.png');
+    await barButton(page, 'Remove from collection').click();
+    await expect(toast(page)).toHaveText('Removed 1 image from “Experiments”.');
+    await expect(imageCards(page)).toHaveCount(7);
+});
+
+test('a failed add keeps the selection; a stale source reloads the images', async ({ page, request }) => {
+    await seed(request);
+    await page.goto(`${workspace}#/images`);
+    await expect(imageCards(page)).toHaveCount(21);
+    let reply = { status: 500, body: { success: false, error: 'The collection could not be changed.' } };
+    await page.route(`**${API}/items`, route => route.fulfill({ status: reply.status, contentType: 'application/json',
+        body: JSON.stringify(reply.body) }));
+    await check(page, 'loose.png');
+    await check(page, 'landscape_05.png');
+    await addChecked(page, 'Grids');
+    await expect(toast(page)).toHaveText('The images could not be added. Try again.');
+    await expect(bar(page).locator('strong')).toHaveText('2 selected');
+    await expect(rows(page)).toHaveCount(2);
+
+    reply = { status: 404, body: { success: false, code: 'files-not-found', error: 'x' } };
+    await addChecked(page, 'Grids');
+    await expect(toast(page)).toHaveText('None of the selected images could be found. Refresh the images and try again.');
+    await expect(bar(page).locator('strong')).toHaveText('2 selected');
+
+    reply = { status: 409, body: { success: false, code: 'source-changed', error: 'x' } };
+    const reloaded = page.waitForRequest(request => request.url().includes('/prompt_manager/gallery/subfolders'));
+    await addChecked(page, 'Grids');
+    await expect(toast(page)).toHaveText('The sources changed. Refresh the images and try again.');
+    await reloaded;
+    await expect(imageCards(page)).toHaveCount(21);
+    await expect(bar(page)).toHaveCount(0);
+});
+
+test('more than 500 or an unloaded "all matching" selection is refused with the cap', async ({ page, request }) => {
+    await seed(request);
+    await page.goto(`${workspace}#/images`);
+    await expect(imageCards(page)).toHaveCount(21);
+    // Every loaded image checked and the listing says more match: "Select all N matching" (#392, D135).
+    await page.route('**/prompt_manager/images/output?*', async route => {
+        // A next page never answers, so the listing stays at its loaded 21 of 900.
+        if (new URL(route.request().url()).searchParams.has('cursor')) return;
+        const response = await route.fetch();
+        const data = await response.json();
+        await route.fulfill({ response, json: { ...data, total: 900, has_more: true, next_cursor: 'more' } });
+    });
+    await page.reload();
+    await expect(imageCards(page)).toHaveCount(21);
+    await page.locator('#image-grid .image-open').first().focus();
+    await page.keyboard.press('ControlOrMeta+a');
+    await bar(page).getByRole('button', { name: 'Select all 900 matching', exact: true }).click();
+    await expect(bar(page).locator('strong')).toHaveText('All 900 matching selected');
+    await barButton(page, 'Add to collection').click();
+    await expect(toast(page)).toHaveText('Select at most 500 images.');
+    await expect(addDialog(page)).toHaveCount(0);
+});
+
+for (const theme of ['dark', 'light']) {
+    test(`the Add to collection dialog fits 390x844 ${theme}`, async ({ page, request }) => {
+        await seed(request);
+        await page.setViewportSize({ width: 390, height: 844 });
+        await page.goto(`${workspace}#/images`);
+        await page.getByRole('combobox', { name: 'Theme', exact: true }).selectOption(theme);
+        await check(page, 'loose.png');
+        await expect(barButton(page, 'Add to collection')).toBeVisible();
+        await barButton(page, 'Add to collection').click();
+        await expect(addDialog(page)).toBeVisible();
+        const fits = await addDialog(page).evaluate(dialog => {
+            const box = dialog.getBoundingClientRect();
+            return box.left >= 0 && box.right <= window.innerWidth && dialog.scrollWidth <= dialog.clientWidth;
+        });
+        expect(fits).toBe(true);
+        expect(await page.evaluate(() => document.documentElement.scrollWidth <= window.innerWidth)).toBe(true);
+        await addField(page).fill('Phone');
+        await addDialog(page).getByRole('button', { name: 'Add', exact: true }).click();
+        await expect(toast(page)).toHaveText('Added 1 image to “Phone”.');
+        await expect(row(page, 'Phone')).toBeVisible();
+    });
+}
+
+// #331 D139 (design §8): the shared inspector's Collections… in the Images side panel and in the viewer.
+const panel = page => page.locator('#image-inspector');
+const collectionsDialog = page => page.getByRole('dialog', { name: 'Collections', exact: true });
+const choice = (page, name) => collectionsDialog(page).getByRole('checkbox', { name, exact: true });
+const newCollection = page => collectionsDialog(page).getByRole('textbox', { name: 'New collection', exact: true });
+
+async function membership(request, path) {
+    const [image] = (await (await request.get(`/prompt_manager/images/output?mode=workspace&q=${encodeURIComponent(path)}`)).json()).images;
+    const query = new URLSearchParams({ root_index: String(image.root_index), root_scope: image.root_scope, path: image.path });
+    const response = await request.get(`${API}/membership?${query}`);
+    await expect(response).toBeOK();
+    return (await response.json()).collection_ids;
+}
+
+test('Collections… in the side panel adds, removes and creates for one image (CUJ 6)', async ({ page, request }) => {
+    const manifest = await seed(request);
+    const { Landscapes, Experiments } = manifest.collections;
+    await page.goto(`${workspace}#/images`);
+    // The newest image, landscape_01, is the panel's.
+    await expect(panel(page).getByRole('heading', { level: 2 })).toHaveText('landscape_01.png');
+    const open = panel(page).getByRole('button', { name: 'Collections…', exact: true });
+    await expect(panel(page).locator('.detail-footer > button').last()).toHaveText('Collections…');
+    await open.click();
+    await expect(collectionsDialog(page)).toBeVisible();
+    await expect(choice(page, 'Landscapes')).toBeChecked();
+    await expect(choice(page, 'Experiments')).not.toBeChecked();
+    await expect(collectionsDialog(page).locator('label.tag-suggestion')).toHaveText(['Experiments', 'Landscapes']);
+    await expect(collectionsDialog(page).locator('.dialog-actions button')).toHaveText(['Cancel', 'Save']);
+    await choice(page, 'Landscapes').uncheck();
+    await choice(page, 'Experiments').check();
+    await newCollection(page).fill('Grids');
+    await collectionsDialog(page).getByRole('button', { name: 'Save', exact: true }).click();
+    await expect(collectionsDialog(page)).toHaveCount(0);
+    await expect(toast(page)).toHaveText('Collections updated.');
+    await expect(open).toBeFocused();
+    await expect(row(page, 'Grids').locator('.nav-count')).toHaveText('1');
+    await expect(row(page, 'Landscapes').locator('.nav-count')).toHaveText('11');
+    await expect(row(page, 'Experiments').locator('.nav-count')).toHaveText('9');
+    const grids = (await listed(request)).find(item => item.name === 'Grids');
+    expect((await membership(request, 'landscape_01')).sort()).toEqual([Experiments.id, grids.id].sort());
+    expect(Landscapes.id).not.toBe(grids.id);
+    // The Images page lists the same files; the dialog opened again shows the new state.
+    await expect(imageCards(page)).toHaveCount(21);
+    await open.click();
+    await expect(choice(page, 'Grids')).toBeChecked();
+    await expect(choice(page, 'Experiments')).toBeChecked();
+    await expect(choice(page, 'Landscapes')).not.toBeChecked();
+    // Save without a change sends nothing.
+    let writes = 0;
+    page.on('request', request => { if (request.url().includes(`${API}/items`)) writes++; });
+    await collectionsDialog(page).getByRole('button', { name: 'Save', exact: true }).click();
+    await expect(collectionsDialog(page)).toHaveCount(0);
+    expect(writes).toBe(0);
+});
+
+test('Collections… in the viewer keeps the viewer until it closes (CUJ 7)', async ({ page, request }) => {
+    const manifest = await seed(request);
+    await page.goto(`${workspace}#/collection/${manifest.collections.Landscapes.id}`);
+    await expect(imageCards(page)).toHaveCount(12);
+    await page.getByRole('button', { name: 'Open landscape_01.png', exact: true }).click();
+    const viewer = page.locator('#viewer-dialog[open]');
+    await expect(viewer.locator('#viewer-title')).toContainText('landscape_01.png');
+    // Review L7: no listing request while the viewer is open.
+    let listings = 0;
+    const countListing = request => { if (new URL(request.url()).pathname === '/prompt_manager/images/output') listings++; };
+    page.on('request', countListing);
+    await viewer.locator('#viewer-details').getByRole('button', { name: 'Collections…', exact: true }).click();
+    await expect(choice(page, 'Landscapes')).toBeChecked();
+    // Review L6: focus moves to the first checkbox once listed.
+    await expect(choice(page, 'Experiments')).toBeFocused();
+    await choice(page, 'Landscapes').uncheck();
+    await collectionsDialog(page).getByRole('button', { name: 'Save', exact: true }).click();
+    await expect(toast(page)).toHaveText('Collections updated.');
+    await expect(collectionsDialog(page)).toHaveCount(0);
+    // The viewer stays on the image; the page keeps listing it while the viewer is open.
+    await expect(viewer.locator('#viewer-title')).toContainText('landscape_01.png');
+    await expect(row(page, 'Landscapes').locator('.nav-count')).toHaveText('11');
+    await expect(imageCards(page)).toHaveCount(12);
+    expect(listings, 'no listing request while the viewer is open').toBe(0);
+    page.off('request', countListing);
+    await viewer.getByRole('button', { name: 'Close image viewer', exact: true }).click();
+    await expect(viewer).toHaveCount(0);
+    await expect(imageCards(page)).toHaveCount(11);
+    await expect(card(page, 'collections/landscape_01.png')).toHaveCount(0);
+});
+
+test('a failed membership read says so and Retry reads it again', async ({ page, request }) => {
+    await seed(request);
+    let failing = true;
+    await page.route(`**${API}/membership?*`, route => failing
+        ? route.fulfill({ status: 500, contentType: 'application/json', body: '{"success":false,"error":"Collections could not be read."}' })
+        : route.fallback());
+    await page.goto(`${workspace}#/images`);
+    await panel(page).getByRole('button', { name: 'Collections…', exact: true }).click();
+    await expect(collectionsDialog(page).getByRole('alert').first()).toHaveText('Collections could not be loaded.');
+    await expect(collectionsDialog(page).getByRole('checkbox')).toHaveCount(0);
+    await expect(collectionsDialog(page).getByRole('button', { name: 'Save', exact: true })).toHaveAttribute('aria-disabled', 'true');
+    failing = false;
+    await collectionsDialog(page).getByRole('button', { name: 'Retry', exact: true }).click();
+    await expect(choice(page, 'Landscapes')).toBeChecked();
+    await expect(collectionsDialog(page).getByRole('button', { name: 'Retry', exact: true })).toBeHidden();
+    // A failed write keeps the dialog with the current membership.
+    await page.route(`**${API}/items`, route => route.fulfill({ status: 500, contentType: 'application/json',
+        body: '{"success":false,"error":"The collection could not be changed."}' }));
+    await choice(page, 'Experiments').check();
+    await collectionsDialog(page).getByRole('button', { name: 'Save', exact: true }).click();
+    await expect(collectionsDialog(page).getByRole('alert')).toHaveText('The collections could not be updated. Try again.');
+    await expect(choice(page, 'Experiments')).not.toBeChecked();
+    await expect(choice(page, 'Landscapes')).toBeChecked();
+});
+
+test('a file in no collection with no collections at all can start one', async ({ page }) => {
+    await page.goto(`${workspace}#/images`);
+    await panel(page).getByRole('button', { name: 'Collections…', exact: true }).click();
+    await expect(collectionsDialog(page)).toContainText('No collections yet. Name one below.');
+    await newCollection(page).fill('First');
+    await collectionsDialog(page).getByRole('button', { name: 'Save', exact: true }).click();
+    await expect(toast(page)).toHaveText('Collections updated.');
+    await expect(rows(page)).toHaveCount(1);
+    await expect(row(page, 'First').locator('.nav-count')).toHaveText('1');
+});
+
+for (const theme of ['dark', 'light']) {
+    test(`the Collections dialog fits 390x844 ${theme}`, async ({ page, request }) => {
+        await seed(request);
+        await page.setViewportSize({ width: 390, height: 844 });
+        await page.goto(`${workspace}#/images`);
+        await page.getByRole('combobox', { name: 'Theme', exact: true }).selectOption(theme);
+        await panel(page).getByRole('button', { name: 'Collections…', exact: true }).click();
+        await expect(choice(page, 'Landscapes')).toBeChecked();
+        const fits = await collectionsDialog(page).evaluate(dialog => {
+            const box = dialog.getBoundingClientRect();
+            return box.left >= 0 && box.right <= window.innerWidth && dialog.scrollWidth <= dialog.clientWidth;
+        });
+        expect(fits).toBe(true);
+        expect(await page.evaluate(() => document.documentElement.scrollWidth <= window.innerWidth)).toBe(true);
+    });
+}
+
+// Review M2: on a collection page, a Save that removed the file from the open collection and then failed keeps its
+// dialog and message; the listing is read again only once the dialog closes.
+test('a partly failed Save from the side panel keeps its dialog until it closes', async ({ page, request }) => {
+    const manifest = await seed(request);
+    const { Experiments, Landscapes } = manifest.collections;
+    // experiment_01 is in both collections; Experiments sorts first, so its remove is sent before Landscapes'.
+    const [image] = (await (await request.get('/prompt_manager/images/output?mode=workspace&q=experiment_01')).json()).images;
+    const item = { root_index: image.root_index, root_scope: image.root_scope, path: image.path };
+    await expect(await request.post(`${API}/items`, { data: { collection_id: Landscapes.id, items: [item] } })).toBeOK();
+    await page.route(`**${API}/items`, route => {
+        const body = route.request().postDataJSON();
+        return route.request().method() === 'DELETE' && body.collection_id === Landscapes.id
+            ? route.fulfill({ status: 500, contentType: 'application/json', body: '{"success":false,"error":"x"}' })
+            : route.fallback();
+    });
+    await page.goto(`${workspace}#/collection/${Experiments.id}`);
+    await expect(imageCards(page)).toHaveCount(8);
+    await expect(panel(page).getByRole('heading', { level: 2 })).toHaveText('experiment_01.png');
+    await panel(page).getByRole('button', { name: 'Collections…', exact: true }).click();
+    await expect(choice(page, 'Landscapes')).toBeChecked();
+    await choice(page, 'Experiments').uncheck();
+    await choice(page, 'Landscapes').uncheck();
+    await collectionsDialog(page).getByRole('button', { name: 'Save', exact: true }).click();
+    await expect(collectionsDialog(page).getByRole('alert')).toHaveText('The collections could not be updated. Try again.');
+    // The remove from Experiments went through; the dialog shows it and the page has not reloaded under it.
+    await expect(choice(page, 'Experiments')).not.toBeChecked();
+    await expect(choice(page, 'Landscapes')).toBeChecked();
+    await expect(imageCards(page)).toHaveCount(8);
+    await collectionsDialog(page).getByRole('button', { name: 'Cancel', exact: true }).click();
+    await expect(collectionsDialog(page)).toHaveCount(0);
+    await expect(imageCards(page)).toHaveCount(7);
+    await expect(card(page, 'collections/experiment_01.png')).toHaveCount(0);
+});

tests/browser/workspace-history.spec.mjs

@@ -495,6 +495,25 @@ test('the inline current file names its prompts apart from the frozen run fields
     expect(await exact(dialog(page).getByRole('region',{name:'Positive prompt',exact:true}).locator('p.prompt-text'))).toBe(before);
 });
 
+// #449 (D154): History's inline current file draws tags in the shared labelled sections, named with its context so they
+// stay apart from the frozen run's fields. The server sends History no image_tags facet today (D88), so the read is
+// given one here to prove the shared render in History.
+test('the inline current file labels its image tags after its negative prompt',async({page})=>{
+    await page.route(url=>url.pathname.endsWith('/inspection'),async route=>{const response=await route.fetch();const body=await response.json();
+        body.prompts={positive:{status:'resolved',text:'current file positive',source:'graph-inferred'},negative:{status:'resolved',text:'current file negative',source:'graph-inferred'}};
+        body.image_tags={state:'current',tags:['history_tag']};
+        await route.fulfill({response,json:body});});
+    await open(page,page.fixture.detail_id);await marker(page,page.fixture.runs.shared).click();await expect(dialog(page).locator('[data-receipt]')).toHaveCount(5);
+    await dialog(page).getByRole('button',{name:'Raw',exact:true}).click();
+    await dialog(page).locator('[data-receipt]').first().getByRole('button',{name:'View current file',exact:true}).click();
+    await expect(page.locator('dialog#viewer-dialog')).toBeVisible();await page.keyboard.press('Escape');await expect(dialog(page)).toBeVisible();
+    const current=dialog(page).getByRole('region',{name:'Current reported file',exact:true});
+    const tags=current.getByRole('region',{name:'Current file image tags',exact:true});
+    await expect(tags.locator('.role-label')).toHaveText('Image tags · 1');await expect(tags.locator('.tags > .chip')).toHaveText(['history_tag']);
+    const negative=current.getByRole('region',{name:'Current file negative prompt',exact:true});
+    expect(await negative.evaluate((node,section)=>({after:Boolean(node.compareDocumentPosition(section)&Node.DOCUMENT_POSITION_FOLLOWING),inside:node.contains(section)}),await tags.elementHandle())).toEqual({after:true,inside:false});
+});
+
 // #332 CUJ 5 (MC8, §5.3): History's current-file inspector reads the file's own category with the Images key rule, so
 // the inline current file and its viewer show "Portraits · this image"; current.png has two saved links, so it inherits
 // nothing and shows no chip before.

tests/browser/workspace-images.spec.mjs

@@ -1,6 +1,6 @@
 import { startCoverage, saveCoverage } from '../../tools/vendor/browser-coverage.mjs';
 import { test, expect } from './fixtures.mjs';
-import { seedConcept, openConcept, openProduction } from './concept-inventory.mjs';
+import { seedConcept, openConcept, openProduction, startInGrid } from './concept-inventory.mjs';
 
 const workspace = '/prompt_manager/workspace#/images';
 const cards = page => page.locator('article[data-image-key]');
@@ -814,6 +814,8 @@ test('leaving during Rescan cancels it without replacing the next workspace', as
 
 test('Images uses approved geometry while keeping real controls', async ({page,request}) => {
     await request.post('/__test__/reset',{data:{workspace_visual:true}});
+    // The approved Grid geometry (D157: the default layout is Masonry).
+    await startInGrid(page);
     await page.setViewportSize({width:1600,height:1100});
     await page.goto(workspace);
     await expect(cards(page)).toHaveCount(24);
@@ -895,7 +897,8 @@ test('images toolbar and filter row match the concept controls', async ({ page,
     // app.js toolbar() a347–360: search, category, sort, Save view, Filters.
     await expect(main.locator('search.toolbar > *')).toHaveCount(5);
     await expect(main.locator('search.toolbar > button')).toHaveText(['Save view', 'Filters']);
-    await expect(sortBy(page).locator('option')).toHaveText(['Recently added', 'Name A–Z', 'Highest rated', 'Oldest first', 'Name Z–A']);
+    // D153 (#447): no "Highest rated", even with a rated prompt in the library.
+    await expect(sortBy(page).locator('option')).toHaveText(['Recently added', 'Name A–Z', 'Oldest first', 'Name Z–A']);
     const row = main.locator('.filter-row');
     await expect(row.locator(':scope > *')).toHaveCount(5);
     await expect(row.locator(':scope > .segmented[role="group"][aria-label="Image layout"] > button')).toHaveText(['Grid', 'List', 'Masonry']);
@@ -914,8 +917,8 @@ test('images toolbar and filter row match the concept controls', async ({ page,
     await search(page).fill('');
     await expect(cards(page)).toHaveCount(2);
     await expect(cards(page).first()).toHaveAttribute('data-image-key', '0:shared/same.png');
-    await sortBy(page).selectOption('rating');
-    await expect(cards(page).first()).toHaveAttribute('data-image-key', '1:shared/same.png');
+    await sortBy(page).selectOption('name-desc');
+    await expect(sortBy(page)).toHaveValue('name-desc');
     await sortBy(page).selectOption('newest');
     await page.getByRole('combobox', { name: 'Filter by category', exact: true }).selectOption('Landscapes');
     await expect(cards(page)).toHaveCount(1);
@@ -1791,7 +1794,7 @@ test('card checkboxes select images into the concept bar and Clear empties it',
     await checkbox(first, 'same.png').check();
     await expect(bulkbar(page).locator('.bulkbar > strong')).toHaveText('1 selected');
     // D85: Add tags (the concept's) and Auto-tag selected lead the real actions.
-    await expect(barButtons(page)).toHaveText(['Add tags', 'Auto-tag selected', 'Set category', 'Export metadata', 'Clear']);
+    await expect(barButtons(page)).toHaveText(['Add tags', 'Auto-tag selected', 'Set category', 'Add to collection', 'Export metadata', 'Clear']);
     // Toggling neither opens the viewer, nor inspects, nor moves focus.
     await expect(page.locator('dialog[open]')).toHaveCount(0);
     await expect(checkbox(first, 'same.png')).toBeFocused();

tests/browser/workspace-library-editor.spec.mjs

@@ -1,7 +1,7 @@
 import {writeFile} from 'node:fs/promises';
 import {startCoverage,saveCoverage} from '../../tools/vendor/browser-coverage.mjs';
 import {test,expect} from './fixtures.mjs';
-import {armUnseenRestores,unseenRestore} from './unseen-restore.mjs';
+import {armUnseenRestores,unseenRestore,followRestore,SEEN_RESTORE_PILL,SEEN_RESTORE_TOAST} from './unseen-restore.mjs';
 import {captureCanonicalState} from './history-canonical-comparison.mjs';
 import {inventoryOf,mapConcept,diffInventory} from './library-editor-canonical-comparison.mjs';
 
@@ -9,8 +9,9 @@ import {inventoryOf,mapConcept,diffInventory} from './library-editor-canonical-c
 test.use({locale:'en-US',timezoneId:'UTC',reducedMotion:'reduce'});
 const workspace='/prompt_manager/workspace';
 const ACTIONS={new:['Cancel','Save'],metadata:['Create text revision','Archive…','Cancel','Save'],revision:['Cancel','Save'],negative:['Cancel','Save']};
-const LABELS=['Title','Positive prompt','Negative prompt (optional)','Category','Tags, separated by commas','Rating','Notes'];
-const RATINGS=[['5','5 / 5'],['4','4 / 5'],['3','3 / 5'],['2','2 / 5'],['1','1 / 5'],['0','Unrated']];
+// D153 (#447): no Rating field; the fixture's positive stays stored with rating 4 through every workspace save.
+const LABELS=['Title','Positive prompt','Negative prompt (optional)','Category','Tags, separated by commas','Notes'];
+const UNAPPLIED='Saved as a new negative variant of an existing prompt. The title, category, tags and notes you entered were not applied.';
 const editorDialog=page=>page.locator('dialog.library-editor');
 
 test.beforeEach(async({page,request,browserName})=>{
@@ -65,9 +66,8 @@ test('edit recipe metadata shows the concept fields with the selected saved nega
     await expect(dialog.getByRole('region')).toHaveCount(0);
     await expect(field(dialog,'Category')).toHaveValue('Portraits');
     await expect(field(dialog,'Tags, separated by commas')).toHaveValue('studio, window light');
-    await expect(field(dialog,'Rating')).toHaveValue('4');
     await expect(field(dialog,'Notes')).toHaveValue(f.metadata.notes);
-    expect(await field(dialog,'Rating').locator('option').evaluateAll(nodes=>nodes.map(node=>[node.value,node.textContent]))).toEqual(RATINGS);
+    await expect(dialog.getByLabel('Rating')).toHaveCount(0);await expect(dialog.locator('select, #recipe-rating')).toHaveCount(0);
     expect(await field(dialog,'Category').evaluate(node=>[...node.list.options].map(option=>option.value))).toEqual(['Portraits','Sketches']);
     expect(await field(dialog,'Tags, separated by commas').evaluate(node=>[...node.list.options].map(option=>option.value))).toEqual(['draft,','studio,','window light,']);
     await expect(dialog.locator('[placeholder]')).toHaveCount(0);
@@ -82,19 +82,20 @@ test('editor modes persist through reload with selected saved negative',async({p
     const f=page.fixture;await mountEditor(page);
     let dialog=await openEditor(page,{mode:'metadata',positiveId:f.positive_id,pairId:f.pair_id});await ready(dialog);
     await field(dialog,'Title').fill('Edited portrait');await field(dialog,'Category').fill('');
-    await field(dialog,'Tags, separated by commas').fill('studio, moody');await field(dialog,'Rating').selectOption('0');
+    await field(dialog,'Tags, separated by commas').fill('studio, moody');
     await field(dialog,'Notes').fill('Edited notes');await button(dialog,'Save').click();
     await expect(dialog).toHaveCount(0);
     expect(page.writes.map(write=>`${write.method} ${write.url}`)).toEqual([`PATCH /prompt_manager/library/${f.positive_id}`]);
-    expect(page.writes[0].body.metadata).toEqual({title:'Edited portrait',category:'',tags:['studio','moody'],rating:null,notes:'Edited notes'});
+    expect(page.writes[0].body.metadata).toEqual({title:'Edited portrait',category:'',tags:['studio','moody'],notes:'Edited notes'});
     await expect.poll(()=>page.evaluate(()=>window.editorEvents.reconciled.length)).toBe(1);
     const saved=(await detail(request,f.positive_id)).positive;
-    expect(saved).toMatchObject({title:'Edited portrait',category:null,rating:null,notes:'Edited notes'});expect([...saved.tags].sort()).toEqual(['moody','studio']);
+    // D153: the edit never sends a rating, so the stored 4 stays.
+    expect(f.metadata.rating).toBe(4);expect(saved).toMatchObject({title:'Edited portrait',category:null,rating:4,notes:'Edited notes'});expect([...saved.tags].sort()).toEqual(['moody','studio']);
 
     await page.reload();await mountEditor(page);
     dialog=await openEditor(page,{mode:'metadata',positiveId:f.positive_id,pairId:f.pair_id});await ready(dialog);
     await expect(field(dialog,'Title')).toHaveValue('Edited portrait');await expect(field(dialog,'Category')).toHaveValue('');
-    await expect(field(dialog,'Rating')).toHaveValue('0');await expect(field(dialog,'Negative prompt (optional)')).toHaveValue(f.selected_negative);
+    await expect(field(dialog,'Negative prompt (optional)')).toHaveValue(f.selected_negative);
     await button(dialog,'Create text revision').click();
     await expect(dialog.getByRole('heading',{name:'Create text revision',exact:true})).toBeVisible();
     await expect(editorDialog(page)).toHaveCount(1);expect(await dialog.evaluate(node=>node.open&&node.matches(':modal'))).toBe(true);
@@ -106,12 +107,15 @@ test('editor modes persist through reload with selected saved negative',async({p
     await button(dialog,'Save').click();await expect(dialog).toHaveCount(0);
     const create=page.writes.at(-1);expect(`${create.method} ${create.url}`).toBe('POST /prompt_manager/library');
     expect(create.body).toMatchObject({positive_text:'Editor portrait, revised dusk light',negative_text:f.selected_negative,
-        parent_metadata:{title:'Edited portrait',category:'Uncategorized',rating:null,notes:'Edited notes'}});
+        parent_metadata:{title:'Edited portrait',category:'Uncategorized',notes:'Edited notes'}});
+    expect(Object.hasOwn(create.body.parent_metadata,'rating'),'D153: a revision sends no rating').toBe(false);
     expect([...create.body.parent_metadata.tags].sort()).toEqual(['moody','studio']);
     await expect.poll(()=>page.evaluate(()=>window.editorEvents.reconciled.length)).toBe(1);
     const revision=await page.evaluate(()=>window.editorEvents.committed.at(-1).receipt);
     expect(revision).toMatchObject({positive_created:true,pair_created:true});expect(revision.positive.id).not.toBe(f.positive_id);
-    expect((await detail(request,revision.positive.id)).positive).toMatchObject({title:'Edited portrait',category:'Uncategorized'});
+    // D153: the revision is a new, unrated positive; the revised prompt keeps its stored rating.
+    expect((await detail(request,revision.positive.id)).positive).toMatchObject({title:'Edited portrait',category:'Uncategorized',rating:null});
+    expect((await detail(request,f.positive_id)).positive.rating).toBe(4);
 
     dialog=await openEditor(page,{mode:'new'});await expect(dialog.getByRole('heading',{name:'New prompt',exact:true})).toBeVisible();await ready(dialog);
     expect(await actionNames(dialog)).toEqual(ACTIONS.new);
@@ -120,9 +124,9 @@ test('editor modes persist through reload with selected saved negative',async({p
     await expect(field(dialog,'Title')).toBeFocused();expect(await field(dialog,'Title').evaluate(node=>node.validity.valueMissing)).toBe(true);
     expect(page.writes).toHaveLength(before);
     await field(dialog,'Title').fill('Fresh idea');await field(dialog,'Positive prompt').fill('A fresh lighthouse at dawn');
-    await field(dialog,'Rating').selectOption('5');await button(dialog,'Save').click();await expect(dialog).toHaveCount(0);
+    await button(dialog,'Save').click();await expect(dialog).toHaveCount(0);
     expect(page.writes.at(-1).body).toEqual({expected_epoch:expect.any(String),positive_text:'A fresh lighthouse at dawn',negative_text:'',
-        parent_metadata:{title:'Fresh idea',category:'Uncategorized',tags:[],rating:5,notes:''}});
+        parent_metadata:{title:'Fresh idea',category:'Uncategorized',tags:[],notes:''}});
 
     dialog=await openEditor(page,{mode:'negative',positiveId:f.positive_id});
     await expect(dialog.getByRole('heading',{name:'Add a negative variant',exact:true})).toBeVisible();await ready(dialog);
@@ -134,6 +138,7 @@ test('editor modes persist through reload with selected saved negative',async({p
     const added=await page.evaluate(()=>window.editorEvents.committed.at(-1).receipt);
     expect(added).toMatchObject({positive_created:false,pair_created:true,pair:{negative_text:'grain, dust'}});
     await expect(page.locator('#editor-trigger')).toBeFocused();
+    expect((await detail(request,f.positive_id)).positive.rating,'D153: adding a negative keeps the stored rating').toBe(4);
 });
 
 test('archive confirms the selected pair with fresh counts and writes only on confirm',async({page,request})=>{
@@ -454,6 +459,37 @@ async function leaveAndReturn(page,id) {
 }
 const laneLabel=(page,text)=>variantsPanel(page).locator('.variant-lane').filter({hasText:text});
 
+// #415: a restore the page sees finish while New prompt is open keeps the dialog and its entries; the Library is shown
+// again from the restored data only once the dialog has closed.
+test('a restore seen finishing keeps New prompt open and reloads the Library after it closes',async({page,request})=>{
+    const f=page.fixture;const snapshot=await request.get('/__test__/library-editor/snapshot');await expect(snapshot).toBeOK();
+    const buffer=await snapshot.body();
+    const current=await detail(request,f.positive_id,'active');
+    await expect(await request.patch(`/prompt_manager/library/${f.positive_id}`,{data:{expected_epoch:current.library_epoch,if_match:current.positive.edit_token,metadata:{title:'Retitled elsewhere'}}})).toBeOK();
+    const restoreRun=await followRestore(page);
+    await page.goto(`${workspace}#/library`);const card=page.locator(`article[data-prompt-id="${f.positive_id}"]`);
+    await expect(card).toContainText('Retitled elsewhere');
+    await expect(page.locator('#activity-pill')).toHaveAttribute('href',SEEN_RESTORE_PILL);
+    await headingActions(page).getByRole('button',{name:'+ New prompt',exact:true}).click();
+    const dialog=editorDialog(page);await expect(dialog.getByRole('heading',{name:'New prompt',exact:true})).toBeVisible();await ready(dialog);
+    await field(dialog,'Title').fill('Typed before restore');await field(dialog,'Positive prompt').fill('A prompt typed while storage was restored');
+    await expect(await request.post('/prompt_manager/restore',{multipart:{database_file:{name:'editor.db',mimeType:'application/octet-stream',buffer}}})).toBeOK();
+    page.writes.length=0;
+    await restoreRun.finish();
+    await expect(toast(page)).toHaveText(SEEN_RESTORE_TOAST);
+    await expect(dialog).toBeVisible();
+    await expect(field(dialog,'Title')).toHaveValue('Typed before restore');
+    await expect(field(dialog,'Positive prompt')).toHaveValue('A prompt typed while storage was restored');
+    await expect(card).toContainText('Retitled elsewhere');
+    await button(dialog,'Save').click();
+    await expect(note(dialog)).toHaveText('Prompt storage was restored since this dialog opened. Nothing was saved.');
+    await expect(button(dialog,'Save')).toBeDisabled();expect(page.writes).toHaveLength(1);
+    await button(dialog,'Cancel').click();await expect(dialog).toHaveCount(0);
+    await expect(card).not.toContainText('Retitled elsewhere');await expect(card).toContainText('Editor portrait');
+    await expect(page.locator('#workspace-main h1')).toBeFocused();
+    expect(page.writes).toHaveLength(1);
+});
+
 test('library heading offers Export then + New prompt, and New opens the saved prompt history',async({page,request})=>{
     await page.goto(`${workspace}#/library`);await expect(page.locator('article[data-prompt-id]').first()).toBeVisible();
     const actions=headingActions(page).locator('button');
@@ -773,20 +809,20 @@ test('archived reuse is reported, shown under Show archived and selected read-on
     await uniqueDotNames(page);
     await page.goto(`${workspace}#/library`);await page.getByRole('button',{name:'+ New prompt',exact:true}).click();dialog=editorDialog(page);await ready(dialog);
     await field(dialog,'Title').fill(f.metadata.title);await field(dialog,'Positive prompt').fill(f.positive_text);await field(dialog,'Negative prompt (optional)').fill(f.archived_negative);
-    await field(dialog,'Category').fill(f.metadata.category);await field(dialog,'Tags, separated by commas').fill(f.metadata.tags.join(', '));await field(dialog,'Rating').selectOption(String(f.metadata.rating));await field(dialog,'Notes').fill(f.metadata.notes);
+    await field(dialog,'Category').fill(f.metadata.category);await field(dialog,'Tags, separated by commas').fill(f.metadata.tags.join(', '));await field(dialog,'Notes').fill(f.metadata.notes);
     await button(dialog,'Save').click();await expect(page).toHaveURL(new RegExp(`#/library/${f.positive_id}$`));
     await expect(toast(page)).toHaveText(message);
     await expect(variantsPanel(page).locator('.library-variant-archived .variant-dot')).toHaveAttribute('aria-pressed','true');
     expect(page.writes.every(write=>write.method==='POST')).toBe(true);
 });
 
-test('reuse that leaves parent metadata unapplied discloses it and keeps the draft for Edit metadata',async({page})=>{
+test('reuse that leaves parent metadata unapplied discloses it and keeps the draft for Edit metadata',async({page,request})=>{
     const f=page.fixture;await page.goto(`${workspace}#/library`);
     await page.getByRole('button',{name:'+ New prompt',exact:true}).click();let dialog=editorDialog(page);await ready(dialog);
     await field(dialog,'Title').fill('A different title');await field(dialog,'Positive prompt').fill(f.positive_text);await field(dialog,'Negative prompt (optional)').fill('brand new negative');
     await field(dialog,'Notes').fill('Draft notes');await button(dialog,'Save').click();
     await expect(page).toHaveURL(new RegExp(`#/library/${f.positive_id}$`));
-    await expect(toast(page)).toHaveText('Saved as a new negative variant of an existing prompt. The title, category, tags, rating and notes you entered were not applied.');
+    await expect(toast(page)).toHaveText(UNAPPLIED);
     await expect(page.locator('#workspace-main h1')).toHaveText(f.metadata.title);
     await page.getByRole('button',{name:'Edit metadata',exact:true}).click();dialog=editorDialog(page);await ready(dialog);
     await expect(field(dialog,'Title')).toHaveValue('A different title');await expect(field(dialog,'Notes')).toHaveValue('Draft notes');
@@ -794,6 +830,8 @@ test('reuse that leaves parent metadata unapplied discloses it and keeps the dra
     await page.getByRole('button',{name:'Edit metadata',exact:true}).click();dialog=editorDialog(page);await ready(dialog);
     await expect(field(dialog,'Title')).toHaveValue(f.metadata.title);await button(dialog,'Cancel').click();
     expect(page.writes).toHaveLength(1);
+    // D153: neither the reuse nor the kept draft touches the stored rating.
+    expect((await detail(request,f.positive_id)).positive.rating).toBe(4);
 });
 
 test('reuse with the stored parent metadata entered does not claim it was not applied',async({page})=>{
@@ -801,7 +839,7 @@ test('reuse with the stored parent metadata entered does not claim it was not ap
     await page.getByRole('button',{name:'+ New prompt',exact:true}).click();let dialog=editorDialog(page);await ready(dialog);
     await field(dialog,'Title').fill(f.metadata.title);await field(dialog,'Positive prompt').fill(f.positive_text);await field(dialog,'Negative prompt (optional)').fill('another brand new negative');
     await field(dialog,'Category').fill(f.metadata.category);await field(dialog,'Tags, separated by commas').fill(f.metadata.tags.join(', '));
-    await field(dialog,'Rating').selectOption(String(f.metadata.rating));await field(dialog,'Notes').fill(f.metadata.notes);
+    await field(dialog,'Notes').fill(f.metadata.notes);
     await button(dialog,'Save').click();await expect(page).toHaveURL(new RegExp(`#/library/${f.positive_id}$`));
     await expect(toast(page)).toHaveText('Saved as a new negative variant of an existing prompt.');
     await page.getByRole('button',{name:'Edit metadata',exact:true}).click();dialog=editorDialog(page);await ready(dialog);
@@ -817,13 +855,13 @@ test('reuse of an uncategorized prompt with the category left blank is not unapp
         await page.goto(`${workspace}#/library`);await page.getByRole('button',{name:'+ New prompt',exact:true}).click();const dialog=editorDialog(page);await ready(dialog);
         await field(dialog,'Title').fill(f.metadata.title);await field(dialog,'Positive prompt').fill(f.positive_text);await field(dialog,'Negative prompt (optional)').fill(negative);
         await field(dialog,'Category').fill(category);await field(dialog,'Tags, separated by commas').fill(f.metadata.tags.join(', '));
-        await field(dialog,'Rating').selectOption(String(f.metadata.rating));await field(dialog,'Notes').fill(f.metadata.notes);
+        await field(dialog,'Notes').fill(f.metadata.notes);
         await button(dialog,'Save').click();await expect(page).toHaveURL(new RegExp(`#/library/${f.positive_id}$`));
     };
     await reuse('','blank category negative');
     await expect(toast(page)).toHaveText('Saved as a new negative variant of an existing prompt.');
     await reuse('Landscapes','different category negative');
-    await expect(toast(page)).toHaveText('Saved as a new negative variant of an existing prompt. The title, category, tags, rating and notes you entered were not applied.');
+    await expect(toast(page)).toHaveText(UNAPPLIED);
     expect(page.writes.map(write=>`${write.method} ${write.url}`)).toEqual(['POST /prompt_manager/library','POST /prompt_manager/library']);
 });
 

tests/browser/workspace-library-selection.spec.mjs

@@ -0,0 +1,294 @@
+import {startCoverage,saveCoverage} from '../../tools/vendor/browser-coverage.mjs';
+import {test,expect} from './fixtures.mjs';
+import {armUnseenRestores,unseenRestore} from './unseen-restore.mjs';
+
+// #437 D148: select prompts on the Library shelf and set their category in one step (the legacy admin's Set Category).
+const route='/prompt_manager/workspace#/library';
+const cards=page=>page.locator('article[data-prompt-id]');
+const card=(page,id)=>page.locator(`article[data-prompt-id="${id}"]`);
+const box=(page,id)=>card(page,id).locator('.image-overlays input[type="checkbox"]');
+const bar=page=>page.locator('#library-bulkbar > .bulkbar');
+const dialog=page=>page.getByRole('dialog',{name:'Categorize selected prompts',exact:true});
+const toast=page=>page.locator('.toast');
+const heading=page=>page.locator('#workspace-main h1');
+const tagsDialog=page=>page.getByRole('dialog',{name:'Tag selected prompts',exact:true});
+const chips=(page,id)=>card(page,id).locator('.tags > .chip');
+async function addTags(page,value) {
+    await bar(page).getByRole('button',{name:'Add tags',exact:true}).click();
+    await expect(tagsDialog(page)).toBeVisible();
+    await tagsDialog(page).locator('#action-input').fill(value);
+    await tagsDialog(page).getByRole('button',{name:'Apply',exact:true}).click();
+}
+
+async function shelf(request) {
+    const body=await(await request.get('/prompt_manager/library?limit=100')).json();
+    return body.items;
+}
+/** Every POST to the bulk category route, with its body. */
+function recordWrites(page) {
+    const writes=[];
+    page.on('request',item=>{if(item.method()==='POST'&&new URL(item.url()).pathname==='/prompt_manager/bulk/category')writes.push(item.postDataJSON());});
+    return writes;
+}
+/** The first mounted card's index once two frames in a row agree (the shelf's window follows scrolling a frame later). */
+async function stableFirstIndex(page) {
+    return page.evaluate(()=>new Promise(resolve=>{
+        const first=()=>document.querySelector('article[data-prompt-id]')?.dataset.index;
+        let previous=first();
+        const check=()=>requestAnimationFrame(()=>requestAnimationFrame(()=>{const now=first();if(now===previous)resolve(Number(now));else{previous=now;check();}}));
+        check();
+    }));
+}
+async function setCategory(page,value) {
+    await bar(page).getByRole('button',{name:'Set category',exact:true}).click();
+    await expect(dialog(page)).toBeVisible();
+    await dialog(page).locator('#action-input').fill(value);
+    await dialog(page).getByRole('button',{name:'Apply',exact:true}).click();
+}
+
+test.beforeEach(async({page,request,browserName})=>{
+    await startCoverage(page,browserName);await expect(await request.post('/__test__/reset',{data:{native_library:true}})).toBeOK();
+    page.libraryErrors=[];page.on('pageerror',error=>page.libraryErrors.push(error.message));
+});
+test.afterEach(async({page,browserName})=>{await saveCoverage(page,browserName);expect(page.libraryErrors).toEqual([]);});
+
+test('Set category writes the selected prompts on the server, reloads the shelf and clears the selection',async({page,request})=>{
+    const items=await shelf(request);const [first,second,third]=items;
+    const writes=recordWrites(page);
+    await page.goto(route);await expect(cards(page)).toHaveCount(items.length);
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();
+    // The checkbox is named after the card's title, as the Images overlay names its file.
+    await expect(box(page,first.id)).toHaveAccessibleName(new RegExp(`^Select `));
+    await box(page,first.id).check();await expect(bar(page).locator('strong')).toHaveText('1 selected');
+    await box(page,second.id).check();await expect(bar(page).locator('strong')).toHaveText('2 selected');
+    // Two of three loaded are checked, so Select all N loaded follows the count (D151, #444 ruling).
+    await expect(bar(page).getByRole('button')).toHaveText([`Select all ${items.length} loaded`,'Add tags','Set category','Clear']);
+    await setCategory(page,'  Studio set  ');
+    await expect(toast(page)).toHaveText('Category “Studio set” set on 2 prompts.');
+    await expect(dialog(page)).toHaveCount(0);
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();
+    for(const item of [first,second]) await expect(card(page,item.id).locator('.card-title .chip')).toHaveText('Studio set');
+    await expect(card(page,third.id).locator('.card-title .chip')).toHaveText(third.category || 'Uncategorized');
+    await expect(box(page,first.id)).not.toBeChecked();
+    const after=await shelf(request);
+    expect(after.filter(item=>item.category==='Studio set').map(item=>item.id).sort()).toEqual([first.id,second.id].sort());
+    const epoch=(await(await request.get('/prompt_manager/library/options')).json()).library_epoch;
+    expect(writes).toEqual([{prompt_ids:[first.id,second.id],category:'Studio set',expected_epoch:epoch}]);
+    // The toolbar's category list now offers the new value, and so does the next dialog.
+    await expect(page.locator('#category-filter option',{hasText:'Studio set'})).toHaveCount(1);
+    await box(page,third.id).check();
+    await bar(page).getByRole('button',{name:'Set category',exact:true}).click();
+    await expect(dialog(page).locator('datalist option[value="Studio set"]')).toHaveCount(1);
+    await dialog(page).getByRole('button',{name:'Cancel',exact:true}).click();await expect(dialog(page)).toHaveCount(0);
+    expect(writes).toHaveLength(1);
+});
+
+test('a blank or control-character category keeps the dialog open; Clear and Escape clear the selection',async({page,request})=>{
+    const items=await shelf(request);const writes=recordWrites(page);
+    await page.goto(route);await expect(cards(page)).toHaveCount(items.length);
+    await box(page,items[0].id).check();
+    await bar(page).getByRole('button',{name:'Set category',exact:true}).click();
+    const field=dialog(page).locator('#action-input');
+    await field.fill('   ');await dialog(page).getByRole('button',{name:'Apply',exact:true}).click();
+    await expect(dialog(page)).toBeVisible();await expect(field).toBeFocused();
+    await field.fill('bad\tvalue');await dialog(page).getByRole('button',{name:'Apply',exact:true}).click();
+    await expect(dialog(page)).toBeVisible();
+    expect(await field.evaluate(node=>node.validationMessage)).toBe('A category cannot contain control characters.');
+    await dialog(page).getByRole('button',{name:'Cancel',exact:true}).click();await expect(dialog(page)).toHaveCount(0);
+    expect(writes).toHaveLength(0);
+    await bar(page).getByRole('button',{name:'Clear',exact:true}).click();
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();await expect(heading(page)).toBeFocused();
+    await expect(box(page,items[0].id)).not.toBeChecked();
+    await box(page,items[1].id).check();
+    await bar(page).getByRole('button',{name:'Set category',exact:true}).focus();await page.keyboard.press('Escape');
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();await expect(heading(page)).toBeFocused();
+    // A filter change replaces the shelf and clears the selection too.
+    await box(page,items[1].id).check();await expect(bar(page)).toBeVisible();
+    await page.locator('#search').fill('portrait');
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();
+});
+
+test('Shift+click selects a range across windowed cards, Tab reaches the checkbox at the window edge, and a return keeps the fence',async({page,request,isMobile})=>{
+    await expect(await request.post('/__test__/reset',{data:{native_library:true,extra_prompts:200}})).toBeOK();
+    if(!isMobile)await page.setViewportSize({width:1600,height:1100});
+    const writes=recordWrites(page);
+    await page.goto(route);await expect(cards(page).first()).toBeVisible();
+    const more=page.getByRole('button',{name:'Load more prompts',exact:true});
+    for(let index=0;index<3;index++) {
+        await expect(more).toHaveAttribute('aria-disabled','false');
+        const response=page.waitForResponse(item=>{const url=new URL(item.url());return url.pathname==='/prompt_manager/library'&&url.searchParams.has('cursor');});
+        await more.evaluate(node=>node.click());await response;
+    }
+    await expect(more).toHaveAttribute('aria-disabled','false');
+    const firstId=Number(await cards(page).first().getAttribute('data-prompt-id'));
+    await box(page,firstId).check();
+    // Windowed: fewer cards are mounted than are loaded.
+    expect(await cards(page).count()).toBeLessThan(160);
+    // Tab from the last mounted card's last control lands on the next (unmounted) card's checkbox.
+    const last=cards(page).last();const lastIndex=Number(await last.getAttribute('data-index'));
+    await last.locator('button').last().focus();await page.keyboard.press('Tab');
+    const next=page.locator(`article[data-index="${lastIndex+1}"]`);
+    await expect(next.locator('.image-overlays input')).toBeFocused();
+    // Shift+click that card's checkbox: every card from the first to it is selected, mounted or not.
+    await next.locator('.image-overlays > label').click({modifiers:['Shift']});
+    await expect(bar(page).locator('strong')).toHaveText(`${lastIndex+2} selected`);
+    // Shift+Tab from a checkbox whose previous card is unmounted lands on that card's last control.
+    // Halfway down the loaded cards (short of the continuation, so no page loads), once the window stops moving.
+    await page.evaluate(()=>{const grid=document.querySelector('.library-grid');scrollTo(0,grid.getBoundingClientRect().top+scrollY+grid.offsetHeight/2);});
+    await expect.poll(()=>page.locator('article[data-index="0"]').count()).toBe(0);
+    const topIndex=await stableFirstIndex(page);
+    await page.locator(`article[data-index="${topIndex}"] .image-overlays input`).focus();await page.keyboard.press('Shift+Tab');
+    await expect(page.locator(`article[data-index="${topIndex-1}"] .card-footer .row-actions > button`).last()).toBeFocused();
+    // Leaving for a prompt's history and coming back restores the shelf without a read; its epoch still fences.
+    await page.locator(`article[data-index="${topIndex}"]`).getByRole('button',{name:'Explore history',exact:true}).click();
+    await page.getByRole('button',{name:'← Prompt library',exact:true}).click();
+    await expect(cards(page).first()).toBeVisible();await expect(page.locator('#library-bulkbar')).toBeEmpty();
+    const visible=Number(await cards(page).first().getAttribute('data-prompt-id'));
+    await box(page,visible).check();await setCategory(page,'Returned');
+    await expect(toast(page)).toHaveText('Category “Returned” set on 1 prompt.');
+    const epoch=(await(await request.get('/prompt_manager/library/options')).json()).library_epoch;
+    expect(writes).toEqual([{prompt_ids:[visible],category:'Returned',expected_epoch:epoch}]);
+});
+
+test('a restore the page did not see refuses the write, changes nothing and reloads the shelf',async({page,request})=>{
+    await armUnseenRestores(page);
+    const items=await shelf(request);
+    await page.goto(route);await expect(cards(page)).toHaveCount(items.length);
+    await box(page,items[0].id).check();
+    const backup=await request.get('/prompt_manager/backup');await expect(backup).toBeOK();const buffer=await backup.body();
+    const restore=await unseenRestore(page,()=>request.post('/prompt_manager/restore',{multipart:{database_file:{name:'library.db',mimeType:'application/octet-stream',buffer}}}));
+    await expect(restore).toBeOK();
+    const reload=page.waitForResponse(item=>{const url=new URL(item.url());return url.pathname==='/prompt_manager/library'&&!url.searchParams.has('cursor');});
+    await setCategory(page,'After restore');
+    await expect(toast(page)).toHaveText('Prompt storage was restored since this page loaded. No category was changed.');
+    await reload;await expect(page.locator('#library-bulkbar')).toBeEmpty();
+    expect((await shelf(request)).filter(item=>item.category==='After restore')).toEqual([]);
+    // The reloaded shelf reads the new epoch: the next write goes through.
+    await box(page,items[0].id).check();await setCategory(page,'After reload');
+    await expect(toast(page)).toHaveText('Category “After reload” set on 1 prompt.');
+});
+
+test('an answer that may follow a committed write says so and reloads; a refused write keeps the selection',async({page,request})=>{
+    const items=await shelf(request);
+    await page.goto(route);await expect(cards(page)).toHaveCount(items.length);
+    await box(page,items[0].id).check();
+    await page.route('**/prompt_manager/bulk/category',route=>route.fulfill({status:500,contentType:'application/json',body:'{"success":false,"error":"boom"}'}),{times:1});
+    await setCategory(page,'Refused');
+    await expect(toast(page)).toHaveText('Category could not be set. Try again.');
+    await expect(bar(page).locator('strong')).toHaveText('1 selected');
+    let sent=false;
+    await page.route('**/prompt_manager/bulk/category',async route=>{sent=true;await route.fetch();await route.abort('failed');},{times:1});
+    await setCategory(page,'Unconfirmed');
+    await expect(toast(page)).toHaveText('The category change could not be confirmed. The Library was reloaded.');
+    expect(sent).toBe(true);
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();
+    await expect(card(page,items[0].id).locator('.card-title .chip')).toHaveText('Unconfirmed');
+});
+
+// #442: Add tags writes prompt tags (never image tags) on the selected prompts.
+function recordTagWrites(page) {
+    const writes=[];
+    page.on('request',item=>{if(item.method()==='POST'&&new URL(item.url()).pathname==='/prompt_manager/bulk/tags')writes.push(item.postDataJSON());});
+    return writes;
+}
+async function promptTags(request,id) {
+    return (await shelf(request)).find(item=>item.id===id).tags;
+}
+
+test('Add tags adds prompt tags to the selected prompts, suggests the vocabulary and reloads the shelf',async({page,request})=>{
+    const items=await shelf(request);const [first,second,third]=items;const writes=recordTagWrites(page);
+    const vocabulary=(await(await request.get('/prompt_manager/tags')).json()).tags;
+    await page.goto(route);await expect(cards(page)).toHaveCount(items.length);
+    await box(page,first.id).check();await box(page,second.id).check();
+    await bar(page).getByRole('button',{name:'Add tags',exact:true}).click();
+    await expect(tagsDialog(page).locator('datalist option')).toHaveCount(vocabulary.length);
+    await tagsDialog(page).getByRole('button',{name:'Cancel',exact:true}).click();await expect(tagsDialog(page)).toHaveCount(0);
+    await addTags(page,' studio-lit , Night ,studio-lit');
+    await expect(toast(page)).toHaveText('Added “studio-lit, Night” to 2 prompts.');
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();
+    for(const item of [first,second]) {
+        await expect(chips(page,item.id).filter({hasText:'studio-lit'})).toHaveCount(1);
+        expect(await promptTags(request,item.id)).toEqual(expect.arrayContaining(['studio-lit','Night']));
+    }
+    expect(await promptTags(request,third.id)).not.toContain('studio-lit');
+    const epoch=(await(await request.get('/prompt_manager/library/options')).json()).library_epoch;
+    expect(writes).toEqual([{prompt_ids:[first.id,second.id],tags:['studio-lit','Night'],expected_epoch:epoch}]);
+    // A second add of the same names changes nothing and says so.
+    await box(page,first.id).check();await addTags(page,'Studio-Lit');
+    await expect(toast(page)).toHaveText('1 prompt already had “Studio-Lit”.');
+});
+
+test('the Add tags field states what it refuses, and a restore the page did not see adds nothing',async({page,request})=>{
+    await armUnseenRestores(page);
+    const items=await shelf(request);const writes=recordTagWrites(page);
+    await page.goto(route);await expect(cards(page)).toHaveCount(items.length);
+    await box(page,items[0].id).check();
+    await bar(page).getByRole('button',{name:'Add tags',exact:true}).click();
+    const field=tagsDialog(page).locator('#action-input');
+    for(const [value,message] of [[' , ','Enter at least one tag.'],['x'.repeat(51),'A tag can have at most 50 characters.'],
+        [Array.from({length:21},(_,i)=>`t${i}`).join(','),'Enter at most 20 tags.']]) {
+        await field.fill(value);await tagsDialog(page).getByRole('button',{name:'Apply',exact:true}).click();
+        await expect(tagsDialog(page)).toBeVisible();
+        expect(await field.evaluate(node=>node.validationMessage)).toBe(message);
+    }
+    await tagsDialog(page).getByRole('button',{name:'Cancel',exact:true}).click();
+    expect(writes).toHaveLength(0);
+    const backup=await request.get('/prompt_manager/backup');await expect(backup).toBeOK();const buffer=await backup.body();
+    const restore=await unseenRestore(page,()=>request.post('/prompt_manager/restore',{multipart:{database_file:{name:'library.db',mimeType:'application/octet-stream',buffer}}}));
+    await expect(restore).toBeOK();
+    await addTags(page,'after-restore');
+    await expect(toast(page)).toHaveText('Prompt storage was restored since this page loaded. No tags were added.');
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();
+    expect((await(await request.get('/prompt_manager/tags')).json()).tags).not.toContain('after-restore');
+});
+
+// #444 ruling (D151): select every loaded prompt, as the legacy page's select-all checkbox did for its page.
+test('Ctrl/Cmd+A and Select all N loaded select every loaded prompt, windowed ones included',async({page,request,isMobile})=>{
+    await expect(await request.post('/__test__/reset',{data:{native_library:true,extra_prompts:200}})).toBeOK();
+    if(!isMobile)await page.setViewportSize({width:1600,height:1100});
+    const writes=recordTagWrites(page);
+    await page.goto(route);await expect(cards(page).first()).toBeVisible();
+    const more=page.getByRole('button',{name:'Load more prompts',exact:true});
+    for(let index=0;index<3;index++) {
+        await expect(more).toHaveAttribute('aria-disabled','false');
+        const response=page.waitForResponse(item=>{const url=new URL(item.url());return url.pathname==='/prompt_manager/library'&&url.searchParams.has('cursor');});
+        await more.evaluate(node=>node.click());await response;
+    }
+    await expect(more).toHaveAttribute('aria-disabled','false');
+    const loaded=160;expect(await cards(page).count()).toBeLessThan(loaded);
+    // In the search box Ctrl/Cmd+A keeps its text meaning: it selects the box's text, no prompt.
+    await page.locator('#search').focus();await page.keyboard.press('ControlOrMeta+a');
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();
+    // One checked card offers the rest; the offer selects every loaded prompt and focus moves to the first action.
+    const firstId=Number(await cards(page).first().getAttribute('data-prompt-id'));
+    await box(page,firstId).check();
+    const offer=bar(page).getByRole('button',{name:`Select all ${loaded} loaded`,exact:true});
+    await expect(bar(page).locator(':scope > *').nth(1)).toHaveText(`Select all ${loaded} loaded`);
+    await offer.click();
+    await expect(bar(page).locator('strong')).toHaveText(`${loaded} selected`);
+    await expect(offer).toHaveCount(0);
+    await expect(bar(page).getByRole('button',{name:'Add tags',exact:true})).toBeFocused();
+    // Another batch arrives unchecked: the offer comes back with the new N.
+    {
+        const response=page.waitForResponse(item=>{const url=new URL(item.url());return url.pathname==='/prompt_manager/library'&&url.searchParams.has('cursor');});
+        await more.evaluate(node=>node.click());await response;
+    }
+    await expect(bar(page).getByRole('button',{name:`Select all ${loaded+40} loaded`,exact:true})).toBeVisible();
+    await expect(bar(page).locator('strong')).toHaveText(`${loaded} selected`);
+    // Escape in the grid clears; Ctrl/Cmd+A in the grid selects them all again.
+    const title=cards(page).first().locator('.card-title > button');
+    await title.focus();await page.keyboard.press('Escape');
+    await expect(page.locator('#library-bulkbar')).toBeEmpty();await expect(title).toBeFocused();
+    await page.keyboard.press('ControlOrMeta+a');
+    const all=loaded+40;
+    await expect(bar(page).locator('strong')).toHaveText(`${all} selected`);
+    // While the write runs, every bar action and the offer are aria-disabled.
+    await box(page,firstId).uncheck();
+    let release;const held=new Promise(resolve=>{release=resolve;});
+    await page.route('**/prompt_manager/bulk/tags',async route=>{await held;await route.continue();},{times:1});
+    await addTags(page,'everyone');
+    for(const name of ['Add tags','Set category',`Select all ${all} loaded`]) await expect(bar(page).getByRole('button',{name,exact:true})).toHaveAttribute('aria-disabled','true');
+    release();
+    await expect(toast(page)).toHaveText(`Added “everyone” to ${all-1} prompts.`);
+    expect(writes).toHaveLength(1);expect(new Set(writes[0].prompt_ids).size).toBe(all-1);
+});

tests/browser/workspace-library.spec.mjs

@@ -152,13 +152,25 @@ test('failed library reads retry while late search responses cannot overwrite ne
     await expect(cards(page)).toHaveCount(1);await expect(cards(page).first()).toContainText('Dog landscape');release();await expect(cards(page).first()).toContainText('Dog landscape');
 });
 
+// D153 (#447): a Browse visit that still names the retired "Highest rated" sort opens the shelf on Recently added.
+test('a Library visit with the retired rating sort opens on Recently added',async({page})=>{
+    await page.goto(route);await expect(cards(page)).toHaveCount(3);
+    const sorts=[];page.on('request',item=>{const url=new URL(item.url());if(url.pathname==='/prompt_manager/library')sorts.push(url.searchParams.get('sort'));});
+    await page.goto('/prompt_manager/workspace#/overview');await expect(page.locator('#workspace-main h1')).toHaveText('Overview');
+    await page.evaluate(async()=>{const {browseLibrary}=await import('/prompt_manager/js/workspace/library.mjs');browseLibrary({sort:'rating'});});
+    await expect(cards(page)).toHaveCount(3);
+    await expect(page.getByRole('combobox',{name:'Sort results',exact:true})).toHaveValue('newest');
+    expect(sorts).toEqual(['newest']);
+});
+
 test('library toolbar matches the concept and tags live in Refine this view',async({page})=>{
     await page.goto(route);await expect(cards(page)).toHaveCount(3);
     // app.js toolbar() a347–360: search.toolbar > label.search-field (input#search), select#category-filter, select#sort, Save view, Filters.
     const toolbar=page.locator('#workspace-main search.toolbar');await expect(toolbar).toHaveCount(1);
     expect(await toolbar.evaluate(node=>[...node.children].map(child=>child.tagName.toLowerCase()+(child.id?`#${child.id}`:'')+(child.className?`.${child.className.split(' ').filter(name=>!name.startsWith('ca-')).join('.')}`:'')))).toEqual(['label.search-field','select#category-filter','select#sort','button','button']);
     await expect(toolbar.locator('#search')).toHaveAttribute('placeholder','Search prompts, tags, filenames…');
-    await expect(toolbar.locator('#sort option')).toHaveText(['Recently added','Name A–Z','Highest rated','Oldest first']);
+    // D153 (#447): no "Highest rated".
+    await expect(toolbar.locator('#sort option')).toHaveText(['Recently added','Name A–Z','Oldest first']);
     await expect(toolbar.getByRole('button')).toHaveText(['Save view','Filters']);
     await expect(page.getByRole('button',{name:'Refresh library',exact:true})).toHaveCount(0);
     await expect(page.getByRole('textbox',{name:'Tags',exact:true})).toHaveCount(0);
@@ -239,7 +251,8 @@ test('a sort change atomically discards old row geometry and renders changed par
     } finally {await page.evaluate(()=>window.releaseLibraryFrames());}
     await expect(cards(page)).toHaveCount(3);
     const changed=page.locator(`article[data-prompt-id="${id}"]`);await expect(changed).toContainText('Revised dog landscape');
-    await expect(changed.locator('.card-footer .ca-visually-hidden')).toHaveText('Rated 5 out of 5');await expect(changed.locator('.card-footer > span > [aria-hidden="true"]')).toHaveText('★★★★★');
+    // D153: a stored rating of 5 still draws no stars.
+    await expect(changed.locator('.card-footer > span')).not.toContainText('★');await expect(changed.locator('.card-footer > span > *')).toHaveCount(0);
     await expect(page.getByRole('button',{name:'Retry prompt library',exact:true})).toBeHidden();
 });
 
@@ -445,8 +458,8 @@ test('Library cards clamp long prompts and keep real content inside the card',as
         for(const item of data.items) {
             const card=page.locator(`article[data-prompt-id="${item.id}"]`);
             const footer=card.locator('.card-footer > span');
-            await expect(footer.locator('[aria-hidden="true"]')).toHaveText(item.rating?'★'.repeat(item.rating):[]);
-            await expect(footer.locator('.ca-visually-hidden')).toHaveText(item.rating?`Rated ${item.rating} out of 5`:[]);
+            // D153: no stars or "Rated N out of 5", whatever the stored rating.
+            await expect(footer.locator('> *')).toHaveCount(0);
             await expect(card.locator('.card-content > .tags .chip')).toHaveText(item.tags);
             await expect(card.locator('.card-meta > span').nth(1)).toHaveText(`${item.run_count} run${item.run_count===1?'':'s'}`);
         }
@@ -598,7 +611,7 @@ async function patchShelf(page,patch) {
         await route.fulfill({response,json:body});
     });
 }
-/** The footer line as seen: everything but its visually hidden rating sentence. */
+/** The footer line as seen: everything but visually hidden text (D153: there is no rating sentence any more). */
 const shownText=node=>[...node.childNodes].filter(child=>!child.classList?.contains('ca-visually-hidden')).map(child=>child.textContent).join('');
 const longPrompt=[`${'A luminous alpine observatory with layered glass terraces '.repeat(6)}`,'second paragraph keeps the full text',`final line ${'detail '.repeat(30)}end of prompt`].join('\n');
 
@@ -618,7 +631,8 @@ test('card matches the concept and never wraps the category mid-word',async({pag
         const card=page.locator(`article[data-prompt-id="${cat.id}"]`);
         // app.js renderLibrary a586: article.prompt-card > .card-art, .card-content, .card-footer.
         await expect(card).toHaveClass('prompt-card');
-        expect(await card.evaluate(node=>[...node.children].map(child=>child.className))).toEqual(['card-art','card-content','card-footer']);
+        // D148 (#437): the concept's Images selection overlay leads the card.
+        expect(await card.evaluate(node=>[...node.children].map(child=>child.className))).toEqual(['image-overlays','card-art','card-content','card-footer']);
         expect(await card.locator('.card-art').evaluate(node=>[...node.children].map(child=>child.tagName))).toEqual(['IMG','IMG']);
         expect(await card.locator('.card-content').evaluate(node=>[...node.children].map(child=>child.className))).toEqual(['card-title','prompt-inspector','card-meta','tags']);
         expect(await card.locator('.card-title').evaluate(node=>[...node.children].map(child=>`${child.tagName}.${child.className}`))).toEqual(['BUTTON.','SPAN.chip']);
@@ -633,13 +647,13 @@ test('card matches the concept and never wraps the category mid-word',async({pag
         await expect(card.locator('.tags > .chip')).toHaveText(['cats','studio']);
         const footer=card.locator('.card-footer');
         expect(await footer.evaluate(node=>[...node.children].map(child=>`${child.tagName}.${child.className}`))).toEqual(['SPAN.','DIV.row-actions']);
-        await expect(footer.locator('> span > [aria-hidden="true"]')).toHaveText('★★★★★');
-        await expect(footer.locator('> span > .ca-visually-hidden')).toHaveText('Rated 5 out of 5');
-        expect(await footer.locator('> span').evaluate(shownText)).toBe('★★★★★ · Used today');
+        // D153 (#447): the stored rating 5 draws no stars; the footer reads only the last use.
+        await expect(footer.locator('> span > *')).toHaveCount(0);
+        expect(await footer.locator('> span').evaluate(shownText)).toBe('Used today');
         const actions=footer.locator('.row-actions > button');await expect(actions).toHaveCount(2);
         await expect(actions.first()).toHaveClass('send icon-btn');await expect(actions.first()).toHaveAccessibleName('Send positive prompt to ComfyUI: Cat portrait');
         await expect(actions.nth(1)).toHaveAccessibleName('Explore history');await expect(actions.nth(1).locator('svg')).toHaveCount(1);
-        // Unrated and never used: the used label alone; no negatives still reads as a count.
+        // Never used: the used label alone; no negatives still reads as a count.
         const unrated=page.locator(`article[data-prompt-id="${dual.id}"]`);
         expect(await unrated.locator('.card-footer > span').evaluate(shownText)).toBe('Not used yet');
         await expect(unrated.locator('.card-footer > span > [aria-hidden="true"]')).toHaveCount(0);

tests/browser/workspace-saved-views.spec.mjs

@@ -230,6 +230,31 @@ test('a view stored under older rules still lists and opens with its usable valu
     await expect(promptCards(page)).toHaveCount(1);
 });
 
+// D153 (#447): the workspace no longer offers "Highest rated"; the API still stores such a view as it was saved.
+test('a view saved under Highest rated opens on Recently added on Images and the Library', async ({ page, request }) => {
+    const images = await create(request, { page: 'images', name: 'Rated cats', state: { ...IMAGE_DEFAULTS, q: 'cat', sort: 'rating' } });
+    const library = await create(request, { page: 'library', name: 'Rated dogs', state: { q: 'dog', category: '', sort: 'rating', tags: [] } });
+    expect((await listed(request)).map(view => view.state.sort)).toEqual(['rating', 'rating']);
+    const listings = [];
+    page.on('request', item => {
+        const url = new URL(item.url());
+        if (['/prompt_manager/images/output', '/prompt_manager/library'].includes(url.pathname)) listings.push(url.searchParams.get('sort'));
+    });
+    await page.goto(`${workspace}#/saved/${images.id}`);
+    await expect(heading(page)).toHaveText('Rated cats');
+    await expect(sortBy(page)).toHaveValue('newest');
+    expect(await sortBy(page).locator('option').allTextContents()).not.toContain('Highest rated');
+    await expect(search(page)).toHaveValue('cat');
+    await expect(imageCards(page)).toHaveCount(1);
+    await savedNav(page).getByRole('link', { name: 'Rated dogs', exact: true }).click();
+    await expect(heading(page)).toHaveText('Rated dogs');
+    await expect(sortBy(page)).toHaveValue('newest');
+    await expect(search(page)).toHaveValue('dog');
+    await expect(promptCards(page)).toHaveCount(1);
+    expect(listings.length).toBeGreaterThan(1);
+    expect(listings).not.toContain('rating');
+});
+
 test('leaving a saved view returns Images and the Library to their own last state (#249 M2)', async ({ page, request }) => {
     const dogs = await create(request, { page: 'images', name: 'Dogs', state: { ...IMAGE_DEFAULTS, q: 'dogs' } });
     const cats = await create(request, { page: 'library', name: 'Cats', state: { q: 'cat', category: '', sort: 'newest', tags: [] } });

tests/browser/workspace-system-backups.spec.mjs

@@ -1,6 +1,7 @@
 import { startCoverage, saveCoverage } from '../../tools/vendor/browser-coverage.mjs';
 import { test, expect } from './fixtures.mjs';
 import { seedConcept } from './concept-inventory.mjs';
+import { followRestore, SEEN_RESTORE_PILL, SEEN_RESTORE_TOAST } from './unseen-restore.mjs';
 
 // System › Backup & export (#249 System S9 Task 24) on the S8 backend and the concept-shaped fixture: Back up and its
 // status line (D103), the restore review (D105) and its job, the exports, Import / inspect and the recovery policy
@@ -265,6 +266,24 @@ test('a finished restore shows the open page again from the restored data', asyn
     expect(await libraryTotal(request)).toBe(restoredTotal);
 });
 
+// #415: an open dialog waits out a finished restore, except a review: the plan it staged no longer describes storage,
+// so it closes as dismissed (its stage discarded), as before.
+test('a restore seen finishing elsewhere closes an open restore review and discards its stage', async ({ page, request }) => {
+    const bytes = await backupBytes(request);
+    const restoreRun = await followRestore(page);
+    await openBackups(page);
+    await expect(page.locator('#activity-pill')).toHaveAttribute('href', SEEN_RESTORE_PILL);
+    await chooseBackup(page, bytes);
+    await expect(dialog(page).locator('#modal-title')).toHaveText('Restore preview');
+    expect((await serverJson(request, '/__test__/system/backups')).stages).toBe(1);
+    const discarded = page.waitForResponse(response => response.url().includes(`${PREVIEW_URL}/`) && response.request().method() === 'DELETE');
+    await restoreRun.finish();
+    await expect(toast(page)).toHaveText(SEEN_RESTORE_TOAST);
+    expect((await discarded).status()).toBe(200);
+    await expect(dialog(page)).toHaveCount(0);
+    expect((await serverJson(request, '/__test__/system/backups')).stages).toBe(0);
+});
+
 test('invalid file shows the validator message; nothing is staged', async ({ page, request }) => {
     const before = await serverJson(request, '/__test__/system/digest');
     await openBackups(page);

tests/browser/workspace-system.spec.mjs

@@ -79,7 +79,7 @@ test('overview shows real counts and Not checked', async ({ page, request }) =>
     expect(await rowsOf(layout, 'Workspace preferences')).toEqual([
         ['One display configuration', 'Grid and density settings are shared across views.', 'Appearance'],
         ['2 output sources', 'Monitoring and enrichment share one source list.', 'Manage sources'],
-        ['Statistics', 'Runs and ratings by category, with library totals.', 'Explore'],
+        ['Statistics', 'Runs by category, with library totals.', 'Explore'],
     ]);
     expect(writes).toEqual([]);
 });
@@ -373,13 +373,12 @@ test('statistics dialog lists runs by category and library totals', async ({ pag
     await expect(dialog.locator('#modal-title')).toHaveText('Usage statistics');
     const tables = dialog.locator('#modal-body > table');
     await expect(tables.locator('caption')).toHaveText(['Runs by category', 'Library totals']);
-    await expect(tables.first().locator('thead th')).toHaveText(['Category', 'Runs', 'Avg. rating']);
+    // D153 (#447): no rating column and no rated-images total.
+    await expect(tables.first().locator('thead th')).toHaveText(['Category', 'Runs']);
     const cells = table => table.locator('tbody tr').evaluateAll(rows => rows.map(row => [...row.cells].map(cell => cell.textContent)));
-    expect(await cells(tables.first())).toEqual(statistics.by_category.map(row => [row.category ?? 'Uncategorized',
-        number(row.runs), row.average_rating === null ? 'Not rated' : `${number(row.average_rating)} / 5`]));
+    expect(await cells(tables.first())).toEqual(statistics.by_category.map(row => [row.category ?? 'Uncategorized', number(row.runs)]));
     expect(await cells(tables.nth(1))).toEqual([['Prompts', number(statistics.totals.total_prompts)],
-        ['Tags', number(statistics.totals.total_tags)], ['Images', number(statistics.totals.total_images)],
-        ['Rated images', number(statistics.rated_images)]]);
+        ['Tags', number(statistics.totals.total_tags)], ['Images', number(statistics.totals.total_images)]]);
     await page.keyboard.press('Escape');
     await expect(explore).toBeFocused();
     // A statistics read that fails says so; the Overview stays.
@@ -481,17 +480,19 @@ test('defaults apply to a new visit but not to the current view or a saved view'
         await page.getByRole('group', { name: 'Image layout' }).getByRole('button', { name: 'List', exact: true }).click();
         await expect(page.locator('.image-grid')).toHaveAttribute('data-layout', 'list');
         await openAppearance(page);
-        await expect(page.locator('#default-view')).toHaveValue('grid');
-        await page.locator('#default-view').selectOption('masonry');
+        // #451: a fresh browser's Default layout is Masonry.
+        await expect(page.locator('#default-view')).toHaveValue('masonry');
+        await page.locator('#default-view').selectOption('grid');
         await page.locator('#density').selectOption('compact');
         await expect(statusOf(page)).toHaveText('Unsaved changes.');
         await saveAppearance(page, 'Saved in this browser.');
         const stored = await storedSettings(page);
-        expect([stored.workspaceLayout, stored.workspaceDensity]).toEqual(['masonry', 'compact']);
-        // Back in Images (the page session's last visit): the current view keeps its own layout and size.
+        expect([stored.workspaceLayout, stored.workspaceDensity]).toEqual(['grid', 'compact']);
+        // Back in Images (the page session's last visit): the toolbar's List is the current view's own and stays; the
+        // size the toolbar never changed follows the newer default (#451).
         await page.locator('#primary-nav a[data-route="images"]').click();
         await expect(page.locator('.image-grid')).toHaveAttribute('data-layout', 'list');
-        await expect(page.locator('.image-grid')).toHaveAttribute('data-density', 'comfortable');
+        await expect(page.locator('.image-grid')).toHaveAttribute('data-density', 'compact');
         // The saved view opens as before.
         await page.goto(`${workspace}#/saved/${view.id}`);
         await expect(imageCards(page).first()).toBeVisible();
@@ -501,15 +502,109 @@ test('defaults apply to a new visit but not to the current view or a saved view'
         await openImages(page);
         await page.reload();
         await expect(imageCards(page).first()).toBeVisible();
-        await expect(page.locator('.image-grid')).toHaveAttribute('data-layout', 'masonry');
+        await expect(page.locator('.image-grid')).toHaveAttribute('data-layout', 'grid');
         await expect(page.locator('.image-grid')).toHaveAttribute('data-density', 'compact');
-        await expect(page.getByRole('group', { name: 'Image layout' }).getByRole('button', { name: 'Masonry', exact: true }))
+        await expect(page.getByRole('group', { name: 'Image layout' }).getByRole('button', { name: 'Grid', exact: true }))
             .toHaveAttribute('aria-pressed', 'true');
     } finally {
         await request.delete(`/prompt_manager/saved-views/${view.id}`);
     }
 });
 
+/** Choose a Default layout in System › Appearance and save it. */
+async function saveDefaultLayout(page, value) {
+    await openAppearance(page);
+    await page.locator('#default-view').selectOption(value);
+    await saveAppearance(page, 'Saved in this browser.');
+}
+
+const imageLayoutOf = page => page.locator('#workspace-main .image-grid');
+const layoutButton = (page, name) => page.getByRole('group', { name: 'Image layout' }).getByRole('button', { name, exact: true });
+
+test('a fresh browser opens Images in Masonry (#451)', async ({ page }) => {
+    await openImages(page);
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'masonry');
+    await expect(layoutButton(page, 'Masonry')).toHaveAttribute('aria-pressed', 'true');
+    // Nothing was stored for it: the default is the gallery settings' own.
+    expect(await page.evaluate(() => localStorage.getItem('gallerySettings'))).toBeNull();
+    // A stored explicit choice is kept.
+    await page.evaluate(() => localStorage.setItem('gallerySettings', JSON.stringify({ workspaceLayout: 'grid' })));
+    await page.reload();
+    await expect(imageCards(page).first()).toBeVisible();
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'grid');
+});
+
+test('a Default layout saved later reaches the next Images visit in this tab (#451)', async ({ page }) => {
+    // The visit starts from Grid, then System saves Masonry without a reload.
+    await page.addInitScript(() => { if (!localStorage.getItem('gallerySettings')) localStorage.setItem('gallerySettings', '{"workspaceLayout":"grid"}'); });
+    await openImages(page);
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'grid');
+    await page.locator('#system-nav a[data-route="system"]').click();
+    await expect(page.locator('#workspace-main .settings-layout')).toBeVisible();
+    await page.locator('#workspace-main .section-nav').getByRole('button', { name: 'Appearance', exact: true }).click();
+    await expect(page.locator('#timeout')).toBeEnabled();
+    await page.locator('#default-view').selectOption('masonry');
+    await saveAppearance(page, 'Saved in this browser.');
+    await page.locator('#primary-nav a[data-route="images"]').click();
+    await expect(imageCards(page).first()).toBeVisible();
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'masonry');
+    await expect(layoutButton(page, 'Masonry')).toHaveAttribute('aria-pressed', 'true');
+});
+
+test('a Default layout saved in another tab reaches this tab\'s next Images visit (#451)', async ({ page, context }) => {
+    await page.addInitScript(() => { if (!localStorage.getItem('gallerySettings')) localStorage.setItem('gallerySettings', '{"workspaceLayout":"grid"}'); });
+    await openImages(page);
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'grid');
+    const other = await context.newPage();
+    try {
+        await saveDefaultLayout(other, 'list');
+    } finally {
+        await other.close();
+    }
+    await page.locator('#primary-nav a[data-route="library"]').click();
+    await expect(page.locator('#workspace-main .prompt-card').first()).toBeVisible();
+    await page.locator('#primary-nav a[data-route="images"]').click();
+    await expect(imageCards(page).first()).toBeVisible();
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'list');
+    // A toolbar choice afterwards is this view's own again: a return keeps it.
+    await layoutButton(page, 'Masonry').click();
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'masonry');
+    await page.locator('#primary-nav a[data-route="library"]').click();
+    await expect(page.locator('#workspace-main .prompt-card').first()).toBeVisible();
+    await page.locator('#primary-nav a[data-route="images"]').click();
+    await expect(imageCards(page).first()).toBeVisible();
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'masonry');
+});
+
+// #451 D157: a cover crop keeps a portrait image's top; Masonry shows the whole image and keeps the centre.
+test('cover crops keep the top of a portrait image; Masonry stays centred (#451)', async ({ page }) => {
+    const position = locator => locator.evaluate(node => getComputedStyle(node).objectPosition);
+    const portrait = page.locator('#workspace-main article.media-card .image-open img[width="1024"][height="1536"]').first();
+    await openImages(page);
+    await layoutButton(page, 'Grid').click();
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'grid');
+    await expect(portrait).toBeVisible();
+    expect(await position(portrait)).toBe('50% 20%');
+    await layoutButton(page, 'List').click();
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'list');
+    expect(await position(portrait)).toBe('50% 20%');
+    // The side panel's cover (also History's media inspector).
+    await portrait.locator('xpath=ancestor::article[1]').locator('.image-card-actions button').first().click();
+    await expect(panel(page).locator('.detail-cover')).toBeVisible();
+    expect(await position(panel(page).locator('.detail-cover'))).toBe('50% 20%');
+    await layoutButton(page, 'Masonry').click();
+    await expect(imageLayoutOf(page)).toHaveAttribute('data-layout', 'masonry');
+    expect(await position(portrait)).toBe('50% 50%');
+    // The Library's card art and a prompt history's shared gallery.
+    await page.goto(`${workspace}#/library`);
+    const art = page.locator('#workspace-main .prompt-card .card-art img').first();
+    await expect(art).toBeVisible();
+    expect(await position(art)).toBe('50% 20%');
+    await page.goto(`${workspace}#/library/${page.manifest.prompts[0].positive_id}`);
+    await expect(imageCards(page).first()).toBeVisible();
+    expect(await position(imageCards(page).first().locator('.image-open img'))).toBe('50% 20%');
+});
+
 test('the defaults reach the Images page only; embedded galleries keep their own presentation', async ({ page }) => {
     await openAppearance(page);
     await page.locator('#default-view').selectOption('list');
@@ -787,12 +882,13 @@ test('reset restores defaults without touching other keys', async ({ page }) =>
     await expect(page.locator('#default-view')).toHaveValue('list');
     const reset = page.locator('#workspace-main .operation').filter({ hasText: 'Reset display defaults' }).getByRole('button', { name: 'Reset', exact: true });
     await reset.click();
-    await expect(page.locator('#default-view')).toHaveValue('grid');
+    // #451 D157: the default layout is Masonry.
+    await expect(page.locator('#default-view')).toHaveValue('masonry');
     await expect(page.locator('#video-muted')).toBeChecked();
     await expect(statusOf(page)).toHaveText('Unsaved changes.');
     await saveAppearance(page, 'Saved in this browser.');
     const stored = await storedSettings(page);
-    expect(stored).toEqual({ gridColumns: 5, defaultLimit: 200, thumbnailPromptDismissed: 'kept', workspaceLayout: 'grid',
+    expect(stored).toEqual({ gridColumns: 5, defaultLimit: 200, thumbnailPromptDismissed: 'kept', workspaceLayout: 'masonry',
         videoMute: true, showImageInfo: true });
 });
 

tests/browser/workspace-tagging-bulk.spec.mjs

@@ -16,7 +16,7 @@ const toasts = page => page.evaluate(() => window.__toasts);
 const chips = page => modal(page).locator('#tag-suggestions .chip.tag-suggestion');
 const actionButtons = page => modal(page).locator('.dialog-actions button');
 const panel = page => page.getByRole('complementary', { name: 'Selected image details', exact: true });
-const tagRow = page => panel(page).locator('.detail-inner > .tags').last();
+const imageTags = page => panel(page).getByRole('region', { name: 'Image tags', exact: true }).locator('.tags');
 const NOT_DOWNLOADED = 'WD14 SwinV2 is not downloaded. Download it in System › Auto-tagging.';
 
 test.beforeEach(async ({ page, request, browserName }) => {
@@ -73,7 +73,7 @@ const jobStart = page => page.waitForRequest(item => item.method() === 'POST' &&
 
 test('selection bar shows Add tags then Auto-tag selected before Set category', async ({ page }) => {
     await select(page, FIRST);
-    await expect(bar(page).locator(':scope > *')).toHaveText(['1 selected', 'Add tags', 'Auto-tag selected', 'Set category', 'Export metadata', 'Clear']);
+    await expect(bar(page).locator(':scope > *')).toHaveText(['1 selected', 'Add tags', 'Auto-tag selected', 'Set category', 'Add to collection', 'Export metadata', 'Clear']);
 });
 
 test('Add tags runs a job and the cards show the tags after it finishes', async ({ page, request }) => {
@@ -110,7 +110,7 @@ test('Add tags runs a job and the cards show the tags after it finishes', async
     await expect(bar(page).locator('strong')).toHaveText('2 selected');
     expect((await ended(request, job.id)).counters).toEqual({ tagged: 2, unchanged: 0, skipped: 0, errors: 0 });
     // The finished job's files are read again: the selected card's panel shows its new image tags.
-    await expect(tagRow(page).locator('button.chip')).toHaveText(['sunset', 'blue_hour']);
+    await expect(imageTags(page).locator('button.chip')).toHaveText(['sunset', 'blue_hour']);
     expect(await savedTags(request, SECOND)).toEqual(['sunset', 'blue_hour']);
     expect(await savedTags(request, UNLINKED)).toEqual([]);
 });
@@ -275,7 +275,7 @@ test('Review each image walks the selection one dialog at a time and only approv
     expect(await savedTags(request, SECOND)).toEqual([]);
     expect(await savedTags(request, UNLINKED)).toEqual(['sandstone', 'desert']);
     // The reviewed file's panel shows its saved tags without reselecting.
-    await expect(tagRow(page).locator('button.chip')).toHaveText(['mine', 'sandstone']);
+    await expect(imageTags(page).locator('button.chip')).toHaveText(['mine', 'sandstone']);
     await expect.poll(async () => (await (await request.get('/__test__/tagging/captures')).json()).captures).toBe(0);
 });
 

tests/browser/workspace-tagging.spec.mjs

@@ -458,27 +458,32 @@ test('virtualized cards release listeners on unmount', async ({ page, request })
     expect(await count()).toBe(mounted);
 });
 
-// Task C3 (D88): the side panel and the viewer details draw the file's own current image tags first, then D21's linked
-// tags not already shown (case-insensitive). Panel image tags are Tags filter buttons; the viewer's chips are plain.
+// Task C3 (D88), #449 (D154): the side panel and the viewer details draw the file's own current image tags in their own
+// "Image tags" section, then D21's linked tags in a "Prompt tags" or "Recipe tags" section after it. Panel image tags are
+// Tags filter buttons; the viewer's chips are plain.
 const panel = page => page.getByRole('complementary', { name: 'Selected image details', exact: true });
-const tagRow = page => panel(page).locator('.detail-inner > .tags').last();
+const imageTags = page => panel(page).getByRole('region', { name: 'Image tags', exact: true }).locator('.tags');
+const linkedTags = page => panel(page).locator('.tag-section.linked-tags > .tags');
 const inspected = (page, path) => page.waitForResponse(item => new URL(item.url()).pathname === `/prompt_manager/images/inspect/${path}`);
 const infoButton = (page, path) => card(page, path).getByRole('button', { name: `Inspect ${path.split('/').pop()}`, exact: true });
 
 test('inspector shows saved image tags after Save without reselecting', async ({ page }) => {
     await page.goto(images);
     await expect(card(page, FIRST)).toHaveClass(/\bselected\b/);
-    // D21/D39: the linked recipe's own tag, plain.
-    await expect(tagRow(page).locator('.chip')).toHaveText(manifest.pair_tags);
-    await expect(tagRow(page).locator('button.chip')).toHaveCount(0);
+    // D21/D39: the linked recipe's own tag, plain, and no image tags yet.
+    await expect(linkedTags(page).locator('.chip')).toHaveText(manifest.pair_tags);
+    await expect(linkedTags(page).locator('button.chip')).toHaveCount(0);
+    await expect(panel(page).locator('.tag-section.image-tags')).toHaveCount(0);
     await openTag(page, FIRST, { goto: false });
     await manual(page).fill('dunes_tag, Editorial');
     const reread = inspected(page, FIRST);
     await dialog(page).getByRole('button', { name: 'Save manual tags', exact: true }).click();
     expect((await (await reread).json()).image_tags).toEqual({ state: 'current', tags: ['dunes_tag', 'Editorial'] });
-    // The image tags lead as filter buttons; the recipe's "editorial" is already shown as "Editorial".
-    await expect(tagRow(page).locator('.chip')).toHaveText(['dunes_tag', 'Editorial']);
-    await expect(tagRow(page).locator('button.chip')).toHaveText(['dunes_tag', 'Editorial']);
+    // The image tags lead as filter buttons in their own section; the recipe keeps its "editorial" in its own (D154).
+    await expect(imageTags(page).locator('.chip')).toHaveText(['dunes_tag', 'Editorial']);
+    await expect(imageTags(page).locator('button.chip')).toHaveText(['dunes_tag', 'Editorial']);
+    await expect(panel(page).locator('.tag-section.image-tags .role-label')).toHaveText('Image tags · 2');
+    await expect(linkedTags(page).locator('.chip')).toHaveText(manifest.pair_tags);
     await expect(card(page, FIRST)).toHaveClass(/\bselected\b/);
     await expect(panel(page).locator('.panel-heading h2')).toHaveText(manifest.title);
 });
@@ -490,9 +495,10 @@ test('clicking an image tag chip filters Images to the tagged unlinked file', as
     await infoButton(page, UNLINKED).click();
     await read;
     await expect(panel(page).locator('.panel-heading h2')).toHaveText('unlinked.png');
-    await expect(tagRow(page).locator('.chip')).toHaveText(['loose_tag']);
+    await expect(imageTags(page).locator('.chip')).toHaveText(['loose_tag']);
+    await expect(panel(page).locator('.tag-section.linked-tags')).toHaveCount(0);
     const listing = page.waitForRequest(item => item.url().includes('/images/output?') && new URL(item.url()).searchParams.get('tags') === 'loose_tag');
-    await tagRow(page).getByRole('button', { name: 'loose_tag', exact: true }).click();
+    await imageTags(page).getByRole('button', { name: 'loose_tag', exact: true }).click();
     await listing;
     await expect(page.locator('.gallery-body article[data-image-key]')).toHaveCount(1);
     await expect(card(page, UNLINKED)).toBeVisible();
@@ -502,18 +508,53 @@ test('clicking an image tag chip filters Images to the tagged unlinked file', as
 test('viewer details show the same chips', async ({ page, request }) => {
     await saveImageTags(request, FIRST, ['dunes_tag', 'Editorial', 'sunlit']);
     await page.goto(images);
-    await expect(tagRow(page).locator('button.chip')).toHaveText(['dunes_tag', 'Editorial', 'sunlit']);
+    await expect(imageTags(page).locator('button.chip')).toHaveText(['dunes_tag', 'Editorial', 'sunlit']);
     await card(page, FIRST).locator('.image-open').click();
     const details = page.locator('#viewer-dialog[open] #viewer-details');
     await expect(details.locator('.panel-heading h2')).toHaveText(manifest.title);
-    const row = details.locator('.tags').last();
+    const row = details.getByRole('region', { name: 'Image tags', exact: true }).locator('.tags');
     await expect(row.locator('.chip')).toHaveText(['dunes_tag', 'Editorial', 'sunlit']);
     await expect(row.locator('span.chip')).toHaveCount(3);
     await expect(row.locator('button')).toHaveCount(0);
+    await expect(details.locator('.tag-section.linked-tags .chip')).toHaveText(manifest.pair_tags);
     // A file without image tags: the viewer shows only D21's linked tags.
     await page.getByRole('button', { name: 'Next media', exact: true }).click();
     await expect(page.locator('#viewer-dialog[open] #viewer-title')).toHaveText(`${manifest.title} · dunes_00002.png`);
-    await expect(details.locator('.tags').last().locator('.chip')).toHaveText(manifest.pair_tags);
+    await expect(details.locator('.tag-section.linked-tags .chip')).toHaveText(manifest.pair_tags);
+    await expect(details.locator('.tag-section.image-tags')).toHaveCount(0);
+});
+
+// #449 (D154): the image tags are a labelled region of their own after the negative prompt, framed like a prompt section,
+// so they never read as part of it; the same in the side panel and the viewer details.
+async function tagSectionPlacement(root) {
+    return root.evaluate(node => {
+        const negative = node.querySelector('section.prompt-section.negative'), tags = node.querySelector('section.tag-section.image-tags');
+        const style = getComputedStyle(tags), label = getComputedStyle(tags.querySelector('.prompt-heading .role-label'));
+        return {
+            after: Boolean(negative.compareDocumentPosition(tags) & Node.DOCUMENT_POSITION_FOLLOWING), inside: negative.contains(tags),
+            below: tags.getBoundingClientRect().top >= negative.getBoundingClientRect().bottom,
+            border: `${style.borderTopWidth} ${style.borderTopStyle}`, paddingTop: style.paddingTop, marginTop: style.marginTop,
+            label: {transform: label.textTransform, size: label.fontSize, sameColorAsNegative: label.color === getComputedStyle(negative.querySelector('.role-label')).color},
+        };
+    });
+}
+const PLACED = {after: true, inside: false, below: true, border: '1px solid', paddingTop: '15px', marginTop: '15px',
+    label: {transform: 'uppercase', size: '10px', sameColorAsNegative: false}};
+
+test('D154: image tags sit in their own labelled section after the negative prompt, in the panel and the viewer', async ({ page, request }) => {
+    await saveImageTags(request, FIRST, ['dunes_tag', 'sunlit']);
+    await page.goto(images);
+    const section = panel(page).getByRole('region', { name: 'Image tags', exact: true });
+    await expect(section.locator('.role-label')).toHaveText('Image tags · 2');
+    await expect(section.locator('.tags > button.chip')).toHaveText(['dunes_tag', 'sunlit']);
+    expect(await tagSectionPlacement(panel(page))).toEqual(PLACED);
+    await card(page, FIRST).locator('.image-open').click();
+    const details = page.locator('#viewer-dialog[open] #viewer-details');
+    await expect(details.locator('.panel-heading h2')).toHaveText(manifest.title);
+    const viewerSection = details.getByRole('region', { name: 'Image tags', exact: true });
+    await expect(viewerSection.locator('.role-label')).toHaveText('Image tags · 2');
+    await expect(viewerSection.locator('.tags > span.chip')).toHaveText(['dunes_tag', 'sunlit']);
+    expect(await tagSectionPlacement(details)).toEqual(PLACED);
 });
 
 test('search finds an image by its image tag', async ({ page, request }) => {
@@ -570,7 +611,7 @@ const listingOf = (page, check) => page.waitForResponse(item => item.url().inclu
 test('the post-save re-read keeps the panel content until the new read lands', async ({ page }) => {
     await page.goto(images);
     await expect(card(page, FIRST)).toHaveClass(/\bselected\b/);
-    await expect(tagRow(page).locator('.chip')).toHaveText(manifest.pair_tags);
+    await expect(linkedTags(page).locator('.chip')).toHaveText(manifest.pair_tags);
     await page.evaluate(() => {
         const node = document.querySelector('aside[aria-label="Selected image details"]');
         window.__loadingSeen = false;
@@ -589,12 +630,13 @@ test('the post-save re-read keeps the panel content until the new read lands', a
     await expect(toast(page)).toHaveText('1 tags saved for dunes_00001.png.');
     // While the re-read is held, the panel still shows the file's details.
     await expect(panel(page).locator('.panel-heading h2')).toHaveText(manifest.title);
-    await expect(tagRow(page).locator('.chip')).toHaveText(manifest.pair_tags);
+    await expect(linkedTags(page).locator('.chip')).toHaveText(manifest.pair_tags);
     expect(await page.evaluate(() => window.__loadingSeen)).toBe(false);
     const landed = inspected(page, FIRST);
     release();
     await landed;
-    await expect(tagRow(page).locator('.chip')).toHaveText(['held_tag', ...manifest.pair_tags]);
+    await expect(imageTags(page).locator('.chip')).toHaveText(['held_tag']);
+    await expect(linkedTags(page).locator('.chip')).toHaveText(manifest.pair_tags);
     expect(await page.evaluate(() => { window.__loadingWatch.disconnect(); return window.__loadingSeen; })).toBe(false);
     await page.unroute(matcher);
 });
@@ -607,7 +649,7 @@ test('a save under a Tags filter re-reads the listing: the selection stays while
     await infoButton(page, SECOND).click();
     await read;
     const filtered = listingOf(page, query => query.get('tags') === 'shared');
-    await tagRow(page).getByRole('button', { name: 'shared', exact: true }).click();
+    await imageTags(page).getByRole('button', { name: 'shared', exact: true }).click();
     await filtered;
     await expect(cards(page)).toHaveCount(2);
     // The newest match (SECOND) is selected once the filter applies.
@@ -767,7 +809,7 @@ test('a tag save from Tags reloads a filtered Images visit instead of restoring
     await page.goto(images);
     await infoButton(page, SECOND).click();
     const filtered = listingOf(page, query => query.get('tags') === 'shared');
-    await tagRow(page).getByRole('button', { name: 'shared', exact: true }).click();
+    await imageTags(page).getByRole('button', { name: 'shared', exact: true }).click();
     await filtered;
     await expect(cards(page)).toHaveCount(2);
     await expect(card(page, SECOND)).toHaveClass(/\bselected\b/);
@@ -792,7 +834,7 @@ test('a failed listing re-read after a save says so and keeps the cards (M2)', a
     await page.goto(images);
     await infoButton(page, SECOND).click();
     const filtered = listingOf(page, query => query.get('tags') === 'shared');
-    await tagRow(page).getByRole('button', { name: 'shared', exact: true }).click();
+    await imageTags(page).getByRole('button', { name: 'shared', exact: true }).click();
     await filtered;
     await expect(cards(page)).toHaveCount(2);
     await openTag(page, SECOND, { goto: false });

tests/browser/workspace-visual.spec.mjs

@@ -4,6 +4,7 @@ import { createHash } from 'node:crypto';
 import { captureCanonicalState, historyGeometry, compareCanonicalHistory, nativeDialogState, exerciseCurrentPairSend, assertRecordedHistory } from './history-canonical-comparison.mjs';
 import { startCoverage, saveCoverage } from '../../tools/vendor/browser-coverage.mjs';
 import { test, expect } from './fixtures.mjs';
+import { startInGrid } from './concept-inventory.mjs';
 
 test.use({ locale: 'en-US', timezoneId: 'UTC', reducedMotion: 'reduce' });
 const workspace = '/prompt_manager/workspace';
@@ -415,6 +416,8 @@ test('zoomed virtual Images keeps the visible real media mounted during native s
 
 test('supplemental extreme zoom grows real footers preserves the visible anchor and resets reservations', async ({ page, request }) => {
     await request.post('/__test__/reset', { data: { workspace_visual: true } });
+    // The Grid's footer reservations (D157: the default layout is Masonry).
+    await startInGrid(page);
     await page.setViewportSize({ width: 390, height: 844 }); await page.goto(`${workspace}#/images`);
     const cards = page.locator('article[data-image-key]'), grid = page.locator('.image-grid');
     await expect(cards).toHaveCount(24);

tests/browser/workspace-vocabulary.spec.mjs

@@ -1,7 +1,7 @@
 import {startCoverage,saveCoverage} from '../../tools/vendor/browser-coverage.mjs';
 import {writeFile} from 'node:fs/promises';
 import {test,expect} from './fixtures.mjs';
-import {armUnseenRestores,unseenRestore} from './unseen-restore.mjs';
+import {armUnseenRestores,unseenRestore,followRestore,SEEN_RESTORE_PILL,SEEN_RESTORE_TOAST} from './unseen-restore.mjs';
 import {captureCanonicalState} from './history-canonical-comparison.mjs';
 import {pageInventoryOf,inventoryOf,diffInventory,mapVocabularyConcept,mapVocabularyDialog} from './vocabulary-canonical-comparison.mjs';
 
@@ -552,6 +552,31 @@ test('conflict and restore states keep the dialog and write nothing',async({page
     expect(writes.filter(write=>write.url==='/prompt_manager/vocabulary/tags')).toHaveLength(1);
 });
 
+// #415: a restore the page sees finish while Create tag is open keeps the dialog and its entries; the page is shown
+// again from the restored data only once the dialog has closed.
+test('a restore seen finishing keeps an open dialog and reloads the page after it closes',async({page,request})=>{
+    const saved=await backup(request);
+    await expect(await request.put('/prompt_manager/tags/mist',{data:{new_name:'fog'}})).toBeOK();
+    const restoreRun=await followRestore(page);const writes=trackWrites(page);await openPage(page);
+    await expect(page.locator('#activity-pill')).toHaveAttribute('href',SEEN_RESTORE_PILL);
+    await expect(tagRow(page,'fog')).toHaveCount(1);
+    const box=await openDialog(page,main(page).getByRole('button',{name:'+ Create tag',exact:true}),'Create tag');
+    await field(box,'Tag name').fill('sunrise');
+    await expect(await request.post('/prompt_manager/restore',{multipart:{database_file:{name:'vocabulary.db',mimeType:'application/octet-stream',buffer:saved}}})).toBeOK();
+    await restoreRun.finish();
+    await expect(toast(page)).toHaveText(SEEN_RESTORE_TOAST);
+    await expect(box).toBeVisible();await expect(field(box,'Tag name')).toHaveValue('sunrise');
+    await expect(tagRow(page,'fog')).toHaveCount(1);
+    await button(box,'Apply').click();
+    await expect(note(box)).toHaveText('Prompt storage was restored since this dialog opened. Nothing was changed.');
+    expect(await actionNames(box)).toEqual(['Reload','Cancel','Apply']);
+    expect(tagNames(await snapshot(request))).not.toContain('sunrise');
+    await button(box,'Cancel').click();await expect(box).toHaveCount(0);
+    await expect(tagRow(page,'mist')).toHaveCount(1);await expect(tagRow(page,'fog')).toHaveCount(0);
+    await expect(main(page).getByRole('heading',{level:1})).toBeFocused();
+    expect(writes.filter(write=>write.url==='/prompt_manager/vocabulary/tags')).toHaveLength(1);
+});
+
 test('lost response is outcome-unknown',async({page,request})=>{
     const writes=trackWrites(page);await openPage(page);
     await page.route('**/prompt_manager/vocabulary/tags/*',async route=>{

tests/comfyui/execution-history.spec.mjs

@@ -263,15 +263,18 @@ test.describe('public chronological history with the actual acknowledged host br
     });
 });
 
-test('a cached dynamic unknown positive stays readable by run UUID and outside the known raw Library scope', async ({ execution }) => {
-    const text = `unknown positive ${randomUUID()}`, graph = recipeGraph({ text, dynamic: true });
+// #403: the graph cannot show a dynamic positive's raw text; a cached job takes it from the execution it reuses.
+test('a cached dynamic positive takes its raw text from the execution it reuses and joins the Library scope', async ({ execution }) => {
+    const text = `dynamic positive ${randomUUID()}`, graph = recipeGraph({ text, dynamic: true });
     const first = await submit(execution.request, graph);await history(execution.request, first);const known = (await capture(execution, first)).contexts[0];
-    const second = await submit(execution.request, graph);await history(execution.request, second);const unknown = (await capture(execution, second)).contexts[0];
-    expect(unknown.positive.raw).toBeNull();expect(unknown.positive.effective).toBe(known.positive.effective);
+    const second = await submit(execution.request, graph);await history(execution.request, second);const reused = await capture(execution, second);const cached = reused.contexts[0];
+    expect(cached.evidence.positive.effective.source).toBe('host_cache_output');expect(cached.positive.effective).toBe(known.positive.effective);
+    expect(cached.positive.raw).toBe(known.positive.raw);expect(cached.evidence.positive.raw).toMatchObject({ provenance: 'inferred', source: 'graph_semantics' });
+    expect(cached.positive_prompt_id).toBe(known.positive_prompt_id);
     const result = await readExecutionRuns(execution.databasePath, second), runId = result.rows[0].run_id;
-    const detail = await execution.request.get(`/prompt_manager/history/runs/${runId}`);expect(detail.ok()).toBe(true);expect((await detail.json()).contexts[0].positive.raw).toBeNull();
+    const detail = await execution.request.get(`/prompt_manager/history/runs/${runId}`);expect(detail.ok()).toBe(true);expect((await detail.json()).contexts[0].positive.raw).toBe(known.positive.raw);
     const response = await execution.request.get(`/prompt_manager/library/${known.positive_prompt_id}/history`);expect(response.ok()).toBe(true);const window = await response.json();
-    expect(window.recorded_run_count).toBe(1);expect(window.items.map(item => item.run_id)).not.toContain(runId);
+    expect(window.recorded_run_count).toBe(2);expect(window.items.map(item => item.run_id)).toContain(runId);
 });
 
 test('actual PM A,A,B jobs persist two combinations and one identical positive snapshot', async ({ execution }) => {
@@ -369,7 +372,7 @@ test('successful source capture survives downstream failure and real interruptio
     }
 });
 
-test('cached dynamic raw stays unknown while literal affixes and effective cache text survive', async ({ execution }) => {
+test('cached dynamic raw comes from the reused execution while literal affixes and effective cache text survive (#403)', async ({ execution }) => {
     const graph = recipeGraph({ dynamic: true, text: 'runtime dynamic recipe', prepend: ' pre ', append: ' post ' });
     const first = await submit(execution.request, graph);
     await history(execution.request, first);
@@ -377,8 +380,9 @@ test('cached dynamic raw stays unknown while literal affixes and effective cache
     const second = await submit(execution.request, graph);
     await history(execution.request, second);
     const value = (await capture(execution, second)).contexts[0];
-    expect(value.positive).toEqual({ raw: null, prepend: ' pre ', append: ' post ', effective: 'pre runtime dynamic recipe post' });
-    expect(value.evidence.positive.raw.provenance).toBe('unknown');
+    expect(value.positive).toEqual({ raw: 'runtime dynamic recipe', prepend: ' pre ', append: ' post ', effective: 'pre runtime dynamic recipe post' });
+    expect(value.evidence.positive.raw).toMatchObject({ provenance: 'inferred', source: 'graph_semantics' });
+    expect(value.evidence.positive.prepend.source).toBe('graph_literal');
     expect(value.evidence.positive.effective.source).toBe('host_cache_output');
 });
 

tests/comfyui/execution-text-link.spec.mjs

@@ -287,6 +287,14 @@ async function capturedContext(execution, id) {
     return result.captures[result.rows[0].run_id].contexts[0];
 }
 
+// #403: the prompt's history "tracked runs", once every job's capture reached SQLite.
+async function expectTrackedRuns(execution, promptId, expected) {
+    await expect.poll(async () => {
+        const response = await execution.request.get(`/prompt_manager/library/${promptId}/history`);
+        return response.ok() ? (await response.json()).recorded_run_count : null;
+    }, { timeout: 10000, message: 'history tracked runs equal the Library run count' }).toBe(expected);
+}
+
 async function firstRun(execution, page, testCase) {
     const prefix = `textlink/${randomUUID()}`;
     const built = await build(page, { ...testCase, prefix });
@@ -309,6 +317,8 @@ for (const kind of ['PromptManager', 'PromptManagerFocused', 'PromptManagerSplit
         const counts = await runCounts(execution, first.context);
         expect(counts.positive).toBe(3);
         if (pair) expect(counts).toMatchObject({ negative: 0, pair: 3 });
+        // #403: the cached jobs are tracked runs of the prompt too
+        await expectTrackedRuns(execution, first.context.positive_prompt_id, 3);
     });
 }
 
@@ -329,6 +339,7 @@ test.describe('cached runs in one renderer (#347)', () => {
                     else expect(again.cached).toContain(first.promptNode);
                 }
                 expect((await runCounts(execution, first.context)).positive).toBe(3);
+                await expectTrackedRuns(execution, first.context.positive_prompt_id, 3);
             } finally { await setCache(execution, 'RAM_PRESSURE'); }
         });
     }
@@ -352,6 +363,9 @@ test.describe('cached runs in one renderer (#347)', () => {
             const bContext = await capturedContext(execution, second.id);
             expect(bContext.positive.raw).toBe(b);
             expect((await runCounts(execution, bContext)).positive).toBe(1);
+            // #403: the third job reuses A's execution, not the newest (B's)
+            await expectTrackedRuns(execution, first.context.positive_prompt_id, 2);
+            await expectTrackedRuns(execution, bContext.positive_prompt_id, 1);
         } finally { await setCache(execution, 'RAM_PRESSURE'); }
     });
 
@@ -390,10 +404,14 @@ test.describe('cached runs in one renderer (#347)', () => {
         } });
         expect(rewired.cached).toContain(first.promptNode);
         expect((await runCounts(execution, first.context)).positive).toBe(1);
-        // The history capture of a cached job is unchanged: effective text from the cache, no stored-prompt hint
+        // #403: a cached job's history takes the raw text of the execution it reuses and names the prompt
         const cachedContext = await capturedContext(execution, (await queueAgain(execution, first, { seed: 3 })).id);
         expect(cachedContext.positive.effective).toBe(first.context.positive.effective);
-        expect(cachedContext.positive_prompt_id).toBeNull();
+        expect(cachedContext.positive.raw).toBe(first.context.positive.raw);
+        expect(cachedContext.evidence.positive.raw).toMatchObject({ provenance: 'inferred', source: 'graph_semantics' });
+        expect(cachedContext.positive_prompt_id).toBe(first.context.positive_prompt_id);
         expect((await runCounts(execution, first.context)).positive).toBe(2);
+        // The rewired job used the prompt as its negative: neither number counts it
+        await expectTrackedRuns(execution, first.context.positive_prompt_id, 2);
     });
 });

tests/js/gallery-settings.test.js

@@ -64,12 +64,15 @@ test("infinite scrolling is off by default and must be a boolean", () => {
 
 test("the workspace keys have their own defaults and validation", () => {
     const { WORKSPACE_LAYOUTS, WORKSPACE_DENSITIES } = require("../../web/js/gallery-settings.js");
-    assert.equal(DEFAULT_SETTINGS.workspaceLayout, "grid");
+    // #451: Images opens in Masonry unless the browser chose otherwise.
+    assert.equal(DEFAULT_SETTINGS.workspaceLayout, "masonry");
     assert.equal(DEFAULT_SETTINGS.workspaceDensity, "comfortable");
     assert.deepEqual(WORKSPACE_LAYOUTS, ["grid", "list", "masonry"]);
     assert.deepEqual(WORKSPACE_DENSITIES, ["comfortable", "large", "compact"]);
     assert.equal(normalizeGallerySettings({ workspaceLayout: "masonry" }).workspaceLayout, "masonry");
-    assert.equal(normalizeGallerySettings({ workspaceLayout: "mosaic" }).workspaceLayout, "grid");
+    assert.equal(normalizeGallerySettings({ workspaceLayout: "grid" }).workspaceLayout, "grid");
+    assert.equal(normalizeGallerySettings({ workspaceLayout: "mosaic" }).workspaceLayout, "masonry");
+    assert.equal(normalizeGallerySettings(null).workspaceLayout, "masonry");
     assert.equal(normalizeGallerySettings({ workspaceDensity: "large" }).workspaceDensity, "large");
     assert.equal(normalizeGallerySettings({ workspaceDensity: 3 }).workspaceDensity, "comfortable");
     // The legacy gallery's own view mode is separate: it has no masonry.

tests/js/workspace-file-inspector.test.js

@@ -209,16 +209,18 @@ test('a branch named by recorded run history shows the Recorded run badge', asyn
 function tagged(imageTags, library = {status:'linked', tags:['Minimal', 'warm_light', 'editorial'], tags_scope:'positive'}) {
     return {...record(), imageTags, library};
 }
-test('renders current image tags before linked tags without duplicates', async () => {
+test('renders current image tags before linked tags, each without repeats within its source', async () => {
     const {inspectorTagChips} = await load();
     assert.equal(typeof inspectorTagChips, 'function', 'inspectorTagChips is exported');
-    assert.deepEqual(inspectorTagChips(tagged({state:'current', tags:['dunes', 'minimal']}), {filterable:true}),
-        [{name:'dunes', filter:true}, {name:'minimal', filter:true}, {name:'warm_light', filter:true}, {name:'editorial', filter:true}]);
+    // D154 amends D88: a linked tag that is also an image tag stays in its own labelled section.
+    assert.deepEqual(inspectorTagChips(tagged({state:'current', tags:['dunes', 'minimal', 'Dunes']}), {filterable:true}),
+        [{name:'dunes', filter:true, source:'image'}, {name:'minimal', filter:true, source:'image'},
+            {name:'Minimal', filter:true, source:'linked'}, {name:'warm_light', filter:true, source:'linked'}, {name:'editorial', filter:true, source:'linked'}]);
     // D21's rule still decides the linked chips: a recipe's own tags stay plain, an image tag stays a filter.
     assert.deepEqual(inspectorTagChips(tagged({state:'current', tags:['Editorial']}, {status:'linked', tags:['editorial', 'cover'], tags_scope:'pair'}), {filterable:true}),
-        [{name:'Editorial', filter:true}, {name:'cover', filter:false}]);
+        [{name:'Editorial', filter:true, source:'image'}, {name:'editorial', filter:false, source:'linked'}, {name:'cover', filter:false, source:'linked'}]);
     // An unlinked file shows only its own tags; a facet-less record (history, local) only the linked ones.
-    assert.deepEqual(inspectorTagChips(tagged({state:'current', tags:['loose']}, {status:'none', tags:[], tags_scope:null}), {filterable:true}), [{name:'loose', filter:true}]);
+    assert.deepEqual(inspectorTagChips(tagged({state:'current', tags:['loose']}, {status:'none', tags:[], tags_scope:null}), {filterable:true}), [{name:'loose', filter:true, source:'image'}]);
     assert.deepEqual(inspectorTagChips(tagged(null), {filterable:true}).map(chip => chip.name), ['Minimal', 'warm_light', 'editorial']);
     assert.deepEqual(inspectorTagChips(tagged({state:'current', tags:[]}, {status:'ambiguous', tags:['x'], tags_scope:'positive'}), {filterable:true}), []);
 });
@@ -236,6 +238,61 @@ test('none and unavailable render no image chips', async () => {
     }
 });
 
+// #449 (D154): the image's own tags and the linked prompt's or recipe's tags each get a labelled section after the
+// prompt blocks, so per-image tags never read as part of the negative prompt.
+function renderVariant(record, variant, options = {}) {
+    const f = mountedDom(), scope = createScope();
+    const view = mountFileInspector(f.container, {sender:{open() {}}, parentScope:scope, variant, ...options});
+    view.setState({loading:false, error:false, record});
+    return {f, done() {view.dispose(); scope.dispose(); f.restore();}};
+}
+const tagSections = f => f.walk(f.container, n => n.tagName === 'section' && /(^| )tag-section( |$)/.test(n.className ?? ''));
+function sectionSummary(f, section) {
+    const label = f.walk(section, n => /(^| )role-label( |$)/.test(n.className ?? ''))[0];
+    const chips = f.walk(section, n => /(^| )chip( |$)/.test(n.className ?? ''));
+    return {className:section.className, name:section.getAttribute('aria-label'), label:label?.textContent, chips:chips.map(chip => `${chip.tagName}:${chip.textContent}`)};
+}
+test('chips name their source: the image or the linked prompt', async () => {
+    const {inspectorTagChips, tagSectionLabel} = await load();
+    assert.deepEqual(inspectorTagChips(tagged({state:'current', tags:['dunes', 'minimal']}), {filterable:true}).map(chip => chip.source),
+        ['image', 'image', 'linked', 'linked', 'linked']);
+    assert.equal(tagSectionLabel('image', {tags_scope:'pair'}), 'Image tags');
+    assert.equal(tagSectionLabel('linked', {tags_scope:'positive'}), 'Prompt tags');
+    assert.equal(tagSectionLabel('linked', {tags_scope:'pair'}), 'Recipe tags');
+});
+test('image tags and linked tags render in their own labelled sections', async () => {
+    for (const [variant, options, tag] of [['panel', {onTagFilter() {}}, 'button'], ['full', {}, 'span']]) {
+        const {f, done} = renderVariant(tagged({state:'current', tags:['dunes', 'minimal']}), variant, options);
+        try {
+            const sections = tagSections(f).map(section => sectionSummary(f, section));
+            assert.deepEqual(sections, [
+                {className:'tag-section image-tags', name:'Image tags', label:'Image tags · 2', chips:[`${tag}:dunes`, `${tag}:minimal`]},
+                {className:'tag-section linked-tags', name:'Prompt tags', label:'Prompt tags · 3', chips:[`${tag}:Minimal`, `${tag}:warm_light`, `${tag}:editorial`]},
+            ], variant);
+            // No bare chip row follows the prompt blocks any more.
+            const inner = f.container.children.find(n => n.tagName === 'div');
+            assert.equal(inner.children.filter(n => n.className === 'tags').length, 0, variant);
+        } finally { done(); }
+    }
+});
+test('a recipe names its tags and the context names the regions', async () => {
+    const {f, done} = renderVariant(tagged({state:'current', tags:['dunes']}, {status:'linked', tags:['cover'], tags_scope:'pair'}), 'panel', {promptContext:'Current file'});
+    try {
+        assert.deepEqual(tagSections(f).map(section => sectionSummary(f, section)), [
+            {className:'tag-section image-tags', name:'Current file image tags', label:'Image tags · 1', chips:['span:dunes']},
+            {className:'tag-section linked-tags', name:'Current file recipe tags', label:'Recipe tags · 1', chips:['span:cover']},
+        ]);
+    } finally { done(); }
+});
+test('an empty tag source draws no section', async () => {
+    let view = renderVariant(tagged(null), 'full');
+    try { assert.deepEqual(tagSections(view.f).map(section => sectionSummary(view.f, section).label), ['Prompt tags · 3']); } finally { view.done(); }
+    view = renderVariant(tagged({state:'current', tags:['loose']}, {status:'none', tags:[], tags_scope:null}), 'full');
+    try { assert.deepEqual(tagSections(view.f).map(section => sectionSummary(view.f, section).label), ['Image tags · 1']); } finally { view.done(); }
+    view = renderVariant(tagged({state:'none', tags:[]}, {status:'none', tags:[], tags_scope:null}), 'full');
+    try { assert.equal(tagSections(view.f).length, 0); } finally { view.done(); }
+});
+
 // LI-4 (D117): "Copy settings summary" after the six facts of the full inspector consolidates the legacy metadata
 // page's COPY ALL (web/metadata.html copyAllData). The legacy template is read from that page itself and run on the
 // legacy parse, so the summary's text is the legacy page's, line for line, wherever the two agree on the values.
@@ -384,6 +441,285 @@ test('the D117 wrap rule targets only the settings summary button', async () =>
     assert.doesNotMatch(css, /#viewer-details > div > button/);
 });
 
+// #331 D139 (design §8): the shared inspector's "Collections…" footer button and its checklist dialog.
+const collectionsDialog = await import('../../web/js/workspace/inspector-collections.mjs').catch(() => null);
+const dialogModule = () => { assert.ok(collectionsDialog, 'inspector-collections.mjs is not implemented'); return collectionsDialog; };
+const COLLECTION_SCOPE = 'a'.repeat(64);
+const FILE = {root_index:2, root_scope:COLLECTION_SCOPE, path:'a.png'};
+function collectionsStore({member = [3], membership, add, remove} = {}) {
+    const calls = [];
+    const store = {
+        calls,
+        collections:[{id:3, name:'LoRa testing', count:3, missing:0}, {id:5, name:'Portraits', count:1, missing:0}],
+        membership: async target => { calls.push(['membership', target]); return membership ? membership() : [...member]; },
+        add: async (where, items) => { calls.push(['add', where, items]); if (add) return add(where, items);
+            return {collection:{id:where.collectionId ?? 9, name:where.name ?? 'x'}, created:!!where.name, added:1, already:0, notFound:0}; },
+        remove: async (id, items) => { calls.push(['remove', id, items]); if (remove) return remove(id, items);
+            return {collection:{id, name:'x'}, removed:1}; },
+        settled: async () => ({}),
+    };
+    return store;
+}
+function footerButtons(f) {
+    return f.walk(f.container, n => n.tagName === 'button').map(n => n.textContent.trim());
+}
+test('inspector shows Collections… only for a rooted file with a store', async () => {
+    for (const variant of ['panel', 'full']) {
+        const f = mountedDom(), scope = createScope();
+        try {
+            const view = mountFileInspector(f.container, {sender:{open() {}}, parentScope:scope, variant, collections:collectionsStore()});
+            view.setState({loading:false, error:false, record:{...record(), historyTarget:'#/library/1'}});
+            const labels = footerButtons(f);
+            assert.ok(labels.includes('Collections…'), `${variant}: the button is offered`);
+            // After Prompt history, as the last footer action.
+            assert.equal(labels.at(-1), 'Collections…');
+            assert.equal(labels.at(-2), 'Prompt history');
+            const button = f.walk(f.container, n => n.tagName === 'button' && n.textContent.trim() === 'Collections…')[0];
+            assert.equal(button.children[0].tagName, 'svg', 'with the folder icon');
+            view.dispose();
+        } finally { scope.dispose(); f.restore(); }
+    }
+    // No store (Prompt library galleries, Inspect a file) or a file that is not a listed one: no button.
+    const local = await localRecord(localPng('Fox Render.png'));
+    for (const [shown, store] of [[record(), null], [local, collectionsStore()],
+        [{...record(), identity:{...record().identity, root_scope:null}}, collectionsStore()]]) {
+        const f = mountedDom(), scope = createScope();
+        try {
+            const view = mountFileInspector(f.container, {sender:{open() {}}, parentScope:scope, variant:'full', collections:store});
+            view.setState({loading:false, error:false, record:shown});
+            assert.ok(!footerButtons(f).includes('Collections…'));
+            view.dispose();
+        } finally { scope.dispose(); f.restore(); }
+    }
+});
+
+/** Open the Collections dialog over FILE; returns its nodes by role and helpers. */
+function openCollections(store, extra = {}) {
+    const {openCollectionsDialog} = dialogModule();
+    const f = mountedDom(), scope = createScope(), toasts = [], changes = [];
+    let dialog = null, closed = 0;
+    openCollectionsDialog({scope, collections:store, target:FILE, trigger:null, notify:text => toasts.push(text),
+        onChanged:change => changes.push(change), openDialog:args => { dialog = args; f.container.append(...args.nodes); },
+        // As the shell does: closing runs the dialog's onClose once.
+        closeDialog:() => { closed++; dialog?.onClose?.(); }, ...extra});
+    const settle = () => new Promise(resolve => setTimeout(resolve, 0));
+    const find = predicate => f.walk(f.container, predicate);
+    const boxes = () => find(n => n.tagName === 'input' && n.type === 'checkbox');
+    const button = text => find(n => n.tagName === 'button' && n.textContent === text)[0];
+    const text = () => find(n => n.tagName === 'p').map(n => n.textContent).filter(Boolean);
+    return {f, scope, toasts, changes, get dialog() { return dialog; }, get closed() { return closed; }, settle, find, boxes, button, text,
+        done() { scope.dispose(); f.restore(); }};
+}
+
+test('the Collections dialog loads the membership and lists every collection', async () => {
+    const store = collectionsStore();
+    const view = openCollections(store);
+    try {
+        assert.equal(view.dialog.title, 'Collections');
+        assert.ok(view.text().includes('Loading collections…'));
+        assert.equal(view.button('Save').getAttribute('aria-disabled'), 'true');
+        await view.settle();
+        assert.deepEqual(store.calls, [['membership', FILE]]);
+        const labels = view.find(n => n.tagName === 'label' && n.className === 'tag-suggestion');
+        assert.deepEqual(labels.map(label => label.textContent), ['LoRa testing', 'Portraits']);
+        assert.deepEqual(view.boxes().map(box => box.checked), [true, false]);
+        const field = view.find(n => n.tagName === 'input' && n.type !== 'checkbox')[0];
+        assert.equal(field.maxLength, 100);
+        assert.equal(view.find(n => n.tagName === 'label' && n.htmlFor === field.id)[0].textContent, 'New collection');
+        assert.deepEqual(view.find(n => n.className === 'dialog-actions')[0].children.map(n => n.textContent), ['Cancel', 'Save']);
+        assert.equal(view.button('Save').getAttribute('aria-disabled'), null);
+    } finally { view.done(); }
+});
+
+test('without collections the dialog says to name one', async () => {
+    const store = collectionsStore({member:[]});
+    store.collections = [];
+    const view = openCollections(store);
+    try {
+        await view.settle();
+        assert.ok(view.text().includes('No collections yet. Name one below.'));
+        assert.equal(view.boxes().length, 0);
+    } finally { view.done(); }
+});
+
+test('a failed membership read offers Retry', async () => {
+    let failing = true;
+    const store = collectionsStore({membership:() => { if (failing) throw new TypeError('offline'); return [5]; }});
+    const view = openCollections(store);
+    try {
+        await view.settle();
+        assert.ok(view.text().includes('Collections could not be loaded.'));
+        assert.equal(view.button('Retry').hidden, false);
+        assert.equal(view.boxes().length, 0);
+        assert.equal(view.button('Save').getAttribute('aria-disabled'), 'true');
+        failing = false;
+        view.button('Retry').fire('click');
+        await view.settle();
+        assert.deepEqual(view.boxes().map(box => box.checked), [false, true]);
+        assert.equal(view.button('Retry').hidden, true);
+    } finally { view.done(); }
+});
+
+test('Save sends the new name, each added and each removed collection, in that order', async () => {
+    const store = collectionsStore();
+    const view = openCollections(store);
+    try {
+        await view.settle();
+        const [lora, portraits] = view.boxes();
+        lora.checked = false; portraits.checked = true;
+        view.find(n => n.tagName === 'input' && n.type !== 'checkbox')[0].value = '  Grids  ';
+        view.button('Save').fire('click');
+        await view.settle(); await view.settle();
+        assert.deepEqual(store.calls.slice(1), [
+            ['add', {name:'Grids'}, [FILE]],
+            ['add', {collectionId:5}, [FILE]],
+            ['remove', 3, [FILE]],
+        ]);
+        assert.equal(view.closed, 1);
+        assert.deepEqual(view.toasts, ['Collections updated.']);
+        assert.deepEqual(view.changes, [{added:[9, 5], removed:[3]}]);
+    } finally { view.done(); }
+});
+
+test('Save without a change sends nothing and closes', async () => {
+    const store = collectionsStore();
+    const view = openCollections(store);
+    try {
+        await view.settle();
+        view.find(n => n.tagName === 'input' && n.type !== 'checkbox')[0].value = '   ';
+        view.button('Save').fire('click');
+        await view.settle();
+        assert.deepEqual(store.calls.map(call => call[0]), ['membership']);
+        assert.equal(view.closed, 1);
+        assert.deepEqual(view.toasts, []);
+        assert.deepEqual(view.changes, []);
+    } finally { view.done(); }
+});
+
+test('a failed Save keeps the dialog, says so and shows the current membership', async () => {
+    let reads = 0;
+    const store = collectionsStore({membership:() => (reads++ ? [5] : [3]),
+        remove:() => { throw Object.assign(new Error('x'), {status:500}); }});
+    const view = openCollections(store);
+    try {
+        await view.settle();
+        const [lora, portraits] = view.boxes();
+        lora.checked = false; portraits.checked = true;
+        view.button('Save').fire('click');
+        // While it saves, × and Escape keep the dialog and Save is busy.
+        assert.equal(view.dialog.canClose(), false);
+        assert.equal(view.button('Save').getAttribute('aria-disabled'), 'true');
+        for (let index = 0; index < 4; index++) await view.settle();
+        assert.equal(view.closed, 0);
+        assert.deepEqual(view.toasts, []);
+        const alert = view.find(n => n.getAttribute?.('role') === 'alert')[0];
+        assert.equal(alert.textContent, 'The collections could not be updated. Try again.');
+        // The add went through; the list shows what the server holds now. Review M2: the page hears about it only once
+        // the dialog closes, so a listing refresh cannot dispose the dialog before its failure is read.
+        assert.deepEqual(view.changes, []);
+        assert.deepEqual(view.boxes().map(box => box.checked), [false, true]);
+        assert.equal(view.dialog.canClose(), true);
+        assert.equal(view.button('Save').getAttribute('aria-disabled'), null);
+        view.button('Cancel').fire('click');
+        assert.deepEqual(view.changes, [{added:[5], removed:[]}]);
+    } finally { view.done(); }
+});
+
+test('a name the store cannot take is refused on the field, keeping the edits (review L5)', async () => {
+    const store = collectionsStore();
+    const view = openCollections(store);
+    try {
+        await view.settle();
+        const [lora, portraits] = view.boxes();
+        lora.checked = false; portraits.checked = true;
+        const field = view.find(n => n.tagName === 'input' && n.type !== 'checkbox')[0];
+        const validity = [];
+        field.setCustomValidity = text => validity.push(text);
+        field.reportValidity = () => validity.push('reported');
+        field.value = 'bad\u0001name';
+        view.button('Save').fire('click');
+        for (let index = 0; index < 3; index++) await view.settle();
+        assert.deepEqual(store.calls.map(call => call[0]), ['membership'], 'nothing sent and no re-read');
+        assert.deepEqual(validity, ['The collections could not be updated. Try again.', 'reported']);
+        assert.deepEqual(view.boxes().map(box => box.checked), [false, true], 'the checkbox edits stay');
+        assert.equal(view.closed, 0);
+        assert.equal(view.button('Save').getAttribute('aria-disabled'), null);
+        // Editing the field clears its message.
+        field.fire('input');
+        assert.equal(validity.at(-1), '');
+    } finally { view.done(); }
+});
+
+
+test('after a failed Save, Save waits for the membership re-read (review M1)', async () => {
+    let release;
+    const store = collectionsStore({remove:() => { throw new Error('x'); }});
+    store.settled = () => new Promise(resolve => { release = resolve; });
+    const view = openCollections(store);
+    try {
+        await view.settle();
+        view.boxes()[0].checked = false;
+        view.button('Save').fire('click');
+        for (let index = 0; index < 3; index++) await view.settle();
+        assert.equal(typeof release, 'function', 'the store is settling');
+        assert.equal(view.button('Save').getAttribute('aria-disabled'), 'true');
+        assert.ok(view.text().includes('Loading collections…'));
+        const sent = store.calls.length;
+        view.button('Save').fire('click');
+        await view.settle();
+        assert.equal(store.calls.length, sent, 'no stale plan is sent');
+        release({});
+        for (let index = 0; index < 3; index++) await view.settle();
+        assert.equal(view.button('Save').getAttribute('aria-disabled'), null);
+    } finally { view.done(); }
+});
+
+test('a typed name of an existing collection is that collection checkbox (review L4)', async () => {
+    // "portraits" names Portraits (not a member): one add by id, no name request.
+    let store = collectionsStore();
+    let view = openCollections(store);
+    try {
+        await view.settle();
+        view.find(n => n.tagName === 'input' && n.type !== 'checkbox')[0].value = '  PORTRAITS ';
+        view.button('Save').fire('click');
+        for (let index = 0; index < 3; index++) await view.settle();
+        assert.deepEqual(store.calls.slice(1), [['add', {collectionId:5}, [FILE]]]);
+    } finally { view.done(); }
+    // Unchecking LoRa testing while typing its name keeps it: nothing to send.
+    store = collectionsStore();
+    view = openCollections(store);
+    try {
+        await view.settle();
+        view.boxes()[0].checked = false;
+        view.find(n => n.tagName === 'input' && n.type !== 'checkbox')[0].value = 'lora testing';
+        view.button('Save').fire('click');
+        await view.settle();
+        assert.deepEqual(store.calls.map(call => call[0]), ['membership']);
+        assert.equal(view.closed, 1);
+    } finally { view.done(); }
+});
+
+test('focus starts on the field, moves to the first checkbox once listed, and falls back on close (review L6)', async () => {
+    const trigger = {isConnected:false, focus() { assert.fail('a detached trigger is not focused'); }};
+    const fallback = {focused:false, focus() { this.focused = true; }};
+    const view = openCollections(collectionsStore(), {trigger, focusFallback:() => fallback});
+    try {
+        const field = view.find(n => n.tagName === 'input' && n.type !== 'checkbox')[0];
+        assert.equal(view.dialog.focusTarget, field);
+        await view.settle();
+        assert.equal(view.boxes()[0].focused, true);
+        view.button('Cancel').fire('click');
+        assert.equal(fallback.focused, true);
+    } finally { view.done(); }
+    const empty = collectionsStore({member:[]});
+    empty.collections = [];
+    const none = openCollections(empty);
+    try {
+        await none.settle();
+        assert.equal(none.find(n => n.tagName === 'input' && n.type !== 'checkbox')[0].focused, true);
+    } finally { none.done(); }
+});
+
 test('category chip names its origin visibly and shows an own category for any link status (#332 D146)', async () => {
     const base = await localRecord(localPng('a.png'));
     const chipTexts = record => {

tests/js/workspace-images-category.test.js

@@ -59,3 +59,9 @@ test('the toolbar list merges prompt and image categories once each', async () =
     assert.deepEqual(mergeCategories(['A', '', '  ', 7], null), ['A']);
     assert.deepEqual(mergeCategories(undefined, ['landscapes', 'Landscapes']), ['landscapes', 'Landscapes']);
 });
+
+test('the outcome names what was set: images by default, prompts for the Library bar (#437)', async () => {
+    const {setOutcome} = await source;
+    assert.equal(setOutcome({updated:1, unchanged:0, not_found:0}, 'A', 'prompt'), 'Category “A” set on 1 prompt.');
+    assert.equal(setOutcome({updated:2, unchanged:0, not_found:0}, 'A', 'prompt'), 'Category “A” set on 2 prompts.');
+});

tests/js/workspace-images-selection.test.js

@@ -0,0 +1,285 @@
+// The Images selection bar's collection actions (#331 D138, design §7): Add to collection after Set category, Remove
+// from collection beside it while a collection filter is active, the 500 cap and every toast of §7.4.
+import test from 'node:test';
+import assert from 'node:assert/strict';
+
+// Guarded: before the module exists every test fails on an assertion, not on the import.
+const loaded = await import('../../web/js/workspace/images-collections.mjs').catch(() => null);
+const module = () => { assert.ok(loaded, 'images-collections.mjs is not implemented'); return loaded; };
+const { mountSelectionBar } = await import('../../web/js/workspace/images-selection.mjs');
+const { createScope } = await import('../../web/js/shared/toolkit/core.js');
+const strings = await import('../../web/js/workspace/collections.mjs');
+
+const SCOPE = 'b'.repeat(64);
+const image = (path, extra = {}) => ({ root_index: 0, root_scope: SCOPE, path, filename: path.split('/').pop(), ...extra });
+const target = path => ({ root_index: 0, root_scope: SCOPE, path });
+const images = count => Array.from({ length: count }, (_, index) => image(`set/${index}.png`));
+const keyOf = item => `${item.root_index}:${item.path}`;
+const flush = () => new Promise(resolve => setTimeout(resolve, 0));
+const coded = (code, status = 400) => Object.assign(new Error(code), { name: 'ApiError', code, status });
+
+/** A minimal DOM for the selection bar: elements with order, attributes, listeners, focus and a toast. */
+function fakeDOM() {
+    const previous = globalThis.document;
+    let active = null;
+    class Element extends EventTarget {
+        constructor(tag) {
+            super();
+            this.tagName = tag.toUpperCase();
+            this.children = [];
+            this.parentNode = null;
+            this.attributes = {};
+            this.dataset = {};
+            this.className = '';
+            this.id = '';
+            this._text = '';
+        }
+        get textContent() { return this._text + this.children.map(child => child.textContent).join(''); }
+        set textContent(text) { this.children = []; this._text = String(text); }
+        get nextElementSibling() {
+            const siblings = this.parentNode?.children ?? [];
+            return siblings[siblings.indexOf(this) + 1] ?? null;
+        }
+        append(...nodes) {
+            for (const node of nodes) { node.remove(); node.parentNode = this; this.children.push(node); }
+        }
+        after(node) {
+            node.remove();
+            node.parentNode = this.parentNode;
+            this.parentNode.children.splice(this.parentNode.children.indexOf(this) + 1, 0, node);
+        }
+        remove() {
+            if (!this.parentNode) return;
+            this.parentNode.children.splice(this.parentNode.children.indexOf(this), 1);
+            this.parentNode = null;
+        }
+        setAttribute(name, value) { this.attributes[name] = String(value); }
+        getAttribute(name) { return Object.hasOwn(this.attributes, name) ? this.attributes[name] : null; }
+        removeAttribute(name) { delete this.attributes[name]; }
+        querySelector() { return this.children.find(child => child.tagName === 'BUTTON' && child.getAttribute('aria-disabled') !== 'true') ?? null; }
+        querySelectorAll() { return []; }
+        focus() { active = this; }
+        click() { this.dispatchEvent(new Event('click')); }
+    }
+    const toast = new Element('div');
+    const body = new Element('body');
+    globalThis.document = {
+        body,
+        get activeElement() { return active ?? body; },
+        createElement: tag => new Element(tag),
+        createTextNode: text => Object.assign(new Element('#text'), { _text: String(text) }),
+        getElementById: id => (id === 'workspace-toast' ? toast : null),
+    };
+    return { Element, body, toast, restore() { globalThis.document = previous; } };
+}
+
+/** The bar over `items` with `checked` keys; `options` adds the collection props. */
+function mountBar(dom, items, checkedKeys, options = {}) {
+    const scope = createScope();
+    const grid = new dom.Element('div');
+    const focusFallback = new dom.Element('h1');
+    const bar = mountSelectionBar({ scope, api: {}, grid, getItems: () => items, focusFallback, onCategorySet() {},
+        activity: options.activity ?? null, onTagged() {}, getFilter: () => ({}),
+        getListing: () => ({ total: options.total ?? items.length, hasMore: false, positiveCount: null }),
+        checked: Object.freeze(new Set(checkedKeys)), ...options.props });
+    return { scope, bar, labels: () => bar.element.children[0]?.children.map(child => child.textContent) ?? [],
+        button: text => bar.element.children[0].children.find(child => child.textContent === text) };
+}
+
+function store(overrides = {}) {
+    const calls = [];
+    return {
+        calls,
+        collections: [{ id: 3, name: 'Experiments', count: 8, missing: 0 }, { id: 7, name: 'Landscapes', count: 12, missing: 1 }],
+        add: async (where, items) => { calls.push(['add', where, items]); return overrides.add?.(where, items); },
+        remove: async (id, items) => { calls.push(['remove', id, items]); return overrides.remove?.(id, items); },
+    };
+}
+
+test('selection bar lists Add to collection after Set category', () => {
+    const dom = fakeDOM();
+    try {
+        const items = images(2);
+        const collections = store();
+        const plain = mountBar(dom, items, [keyOf(items[0])], { activity: {}, props: { collections } });
+        assert.deepEqual(plain.labels(), ['1 selected', 'Add tags', 'Auto-tag selected', 'Set category', 'Add to collection',
+            'Export metadata', 'Clear']);
+        plain.scope.dispose();
+        // §7.1: Remove from collection beside Add, only while a collection filter is active.
+        const filtered = mountBar(dom, items, [keyOf(items[0])], { activity: {}, props: { collections, collection: '7' } });
+        assert.deepEqual(filtered.labels(), ['1 selected', 'Add tags', 'Auto-tag selected', 'Set category', 'Add to collection',
+            'Remove from collection', 'Export metadata', 'Clear']);
+        filtered.scope.dispose();
+        // Without the client store (a page that has none) neither shows.
+        const none = mountBar(dom, items, [keyOf(items[0])]);
+        assert.deepEqual(none.labels(), ['1 selected', 'Set category', 'Export metadata', 'Clear']);
+        none.scope.dispose();
+    } finally { dom.restore(); }
+});
+
+test('more than 500 checked toasts and opens nothing', async () => {
+    const dom = fakeDOM();
+    try {
+        const items = images(501);
+        const collections = store();
+        const view = mountBar(dom, items, items.map(keyOf), { props: { collections, collection: '7' } });
+        view.button('Add to collection').click();
+        assert.equal(dom.toast.textContent, 'Select at most 500 images.');
+        assert.equal(dom.body.children.length, 0, 'no dialog opened');
+        dom.toast.textContent = '';
+        view.button('Remove from collection').click();
+        await flush();
+        assert.equal(dom.toast.textContent, 'Select at most 500 images.');
+        assert.deepEqual(collections.calls, []);
+        view.scope.dispose();
+    } finally { dom.restore(); }
+});
+
+test('Remove from collection removes the checked files and reloads the listing', async () => {
+    const dom = fakeDOM();
+    try {
+        const items = images(3);
+        let changed = 0;
+        const collections = store({ remove: () => ({ collection: { id: 7, name: 'Landscapes' }, removed: 2 }) });
+        const view = mountBar(dom, items, [keyOf(items[0]), keyOf(items[2])], { props: { collections, collection: '7',
+            onCollectionsChanged: () => { changed++; } } });
+        const remove = view.button('Remove from collection');
+        remove.click();
+        // A second press while the first is in flight is ignored.
+        assert.equal(remove.getAttribute('aria-disabled'), 'true');
+        remove.click();
+        await flush();
+        assert.deepEqual(collections.calls, [['remove', 7, [target('set/0.png'), target('set/2.png')]]]);
+        assert.equal(dom.toast.textContent, 'Removed 2 images from “Landscapes”.');
+        assert.equal(changed, 1);
+        assert.equal(remove.getAttribute('aria-disabled'), null);
+        view.scope.dispose();
+    } finally { dom.restore(); }
+});
+
+test('the collection-target rule keeps the all-matching selection to its loaded images', () => {
+    const { collectionTargets } = module();
+    assert.deepEqual(collectionTargets(images(2)), { targets: [target('set/0.png'), target('set/1.png')] });
+    assert.deepEqual(collectionTargets(images(501)), { refusal: strings.TOO_MANY_TOAST });
+    // All N matching: only when every match is loaded and checked, and at most 500.
+    assert.deepEqual(collectionTargets(images(2), { allMatching: true, total: 2 }), { targets: [target('set/0.png'), target('set/1.png')] });
+    assert.deepEqual(collectionTargets(images(2), { allMatching: true, total: 3 }), { refusal: strings.TOO_MANY_TOAST });
+    assert.deepEqual(collectionTargets(images(501), { allMatching: true, total: 501 }), { refusal: strings.TOO_MANY_TOAST });
+    // A file without a listing identity cannot be named; none left is "could not be found".
+    assert.deepEqual(collectionTargets([{ ...image('x.png'), root_scope: 'nope' }]), { refusal: strings.FILES_NOT_FOUND_TOAST });
+});
+
+const pageScope = () => {
+    const controller = new AbortController();
+    return { signal: controller.signal, cleanup() {}, abort: () => controller.abort() };
+};
+
+test('Add to collection opens the concept dialog with the existing names as suggestions', () => {
+    const { openAddToCollection } = module();
+    let dialog = null;
+    const collections = store();
+    openAddToCollection({ scope: pageScope(), collections, images: images(1), trigger: 'trigger', notify: assert.fail,
+        openInput: args => { dialog = args; } });
+    assert.equal(dialog.title, 'Add selected images to collection');
+    assert.equal(dialog.label, 'Collection name');
+    assert.equal(dialog.action, 'Add');
+    assert.equal(dialog.maxLength, 100);
+    assert.equal(dialog.trigger, 'trigger');
+    assert.deepEqual(dialog.suggestions, ['Experiments', 'Landscapes']);
+    assert.equal(dialog.value ?? '', '');
+});
+
+/** Apply `name` in the Add dialog over `count` images; resolves with the toasts and the store's calls. */
+async function add(name, answer, { count = 1, extra = {} } = {}) {
+    const { openAddToCollection } = module();
+    const toasts = [];
+    let dialog = null, sourceChanged = 0;
+    const collections = store({ add: answer });
+    openAddToCollection({ scope: pageScope(), collections, images: images(count), notify: text => toasts.push(text),
+        openInput: args => { dialog = args; }, onSourceChanged: () => { sourceChanged++; }, ...extra });
+    await dialog.onApply(name);
+    return { toasts, calls: collections.calls, sourceChanged };
+}
+const answer = (added, already, notFound, created = false) => () => ({ collection: { id: 9, name: 'LoRa testing' }, created,
+    added, already, notFound });
+
+test('Add sends the normalized name and the checked files', async () => {
+    const { toasts, calls } = await add('  LoRa   testing ', answer(3, 0, 0, true), { count: 3 });
+    assert.deepEqual(calls, [['add', { name: 'LoRa testing' }, [target('set/0.png'), target('set/1.png'), target('set/2.png')]]]);
+    assert.deepEqual(toasts, ['Added 3 images to “LoRa testing”.']);
+});
+
+test('Add toasts every outcome of §7.4', async () => {
+    const cases = [
+        [answer(1, 0, 0), 'Added 1 image to “LoRa testing”.'],
+        [answer(1, 1, 0), 'Added 1 image to “LoRa testing”. 1 was already in it.'],
+        [answer(2, 3, 0), 'Added 2 images to “LoRa testing”. 3 were already in it.'],
+        [answer(1, 0, 1), 'Added 1 image to “LoRa testing”. 1 could not be found.'],
+        [answer(1, 2, 4), 'Added 1 image to “LoRa testing”. 2 were already in it. 4 could not be found.'],
+        [answer(0, 2, 0), 'All selected images were already in “LoRa testing”.'],
+        [answer(0, 1, 1), 'Added 0 images to “LoRa testing”. 1 was already in it. 1 could not be found.'],
+        [() => { throw coded('files-not-found', 404); }, 'None of the selected images could be found. Refresh the images and try again.'],
+        [() => { throw coded('collection-not-found', 404); }, 'That collection no longer exists.'],
+        [() => { throw new TypeError('offline'); }, 'The images could not be added. Try again.'],
+    ];
+    for (const [reply, text] of cases) assert.deepEqual((await add('LoRa testing', reply)).toasts, [text], text);
+});
+
+test('a stale source on Add says so and reloads the listing', async () => {
+    const { toasts, sourceChanged } = await add('Grids', () => { throw coded('source-changed', 409); });
+    assert.deepEqual(toasts, ['The sources changed. Refresh the images and try again.']);
+    assert.equal(sourceChanged, 1);
+});
+
+test('Add with a name the store cannot take sends nothing and says it failed', async () => {
+    const { toasts, calls } = await add('bad\u0001name', answer(1, 0, 0));
+    assert.deepEqual(calls, []);
+    assert.deepEqual(toasts, ['The images could not be added. Try again.']);
+});
+
+test('Add after the page left shows no toast', async () => {
+    const { openAddToCollection } = module();
+    const scope = pageScope();
+    const toasts = [];
+    let dialog = null;
+    openAddToCollection({ scope, collections: store({ add: answer(1, 0, 0) }), images: images(1), notify: text => toasts.push(text),
+        openInput: args => { dialog = args; } });
+    scope.abort();
+    await dialog.onApply('Grids');
+    assert.deepEqual(toasts, []);
+});
+
+test('Add refuses an over-cap or all-matching selection before any dialog', () => {
+    const { openAddToCollection } = module();
+    const toasts = [];
+    const open = extra => openAddToCollection({ scope: pageScope(), collections: store(), notify: text => toasts.push(text),
+        openInput: () => assert.fail('no dialog opens'), ...extra });
+    open({ images: images(501) });
+    open({ images: images(2), allMatching: true, total: 40 });
+    assert.deepEqual(toasts, [strings.TOO_MANY_TOAST, strings.TOO_MANY_TOAST]);
+});
+
+/** Remove `count` images from collection 7 with `reply`; resolves with the toasts and callbacks. */
+async function remove(reply, count = 1) {
+    const { removeFromCollection } = module();
+    const toasts = [];
+    let changed = 0, sourceChanged = 0;
+    const collections = store({ remove: reply });
+    await removeFromCollection({ scope: pageScope(), collections, collectionId: 7, images: images(count),
+        notify: text => toasts.push(text), onChanged: () => { changed++; }, onSourceChanged: () => { sourceChanged++; } });
+    return { toasts, changed, sourceChanged, calls: collections.calls };
+}
+
+test('Remove toasts every outcome of §7.4', async () => {
+    assert.deepEqual(await remove(() => ({ collection: { id: 7, name: 'Landscapes' }, removed: 1 })),
+        { toasts: ['Removed 1 image from “Landscapes”.'], changed: 1, sourceChanged: 0, calls: [['remove', 7, [target('set/0.png')]]] });
+    assert.deepEqual((await remove(() => ({ collection: { id: 7, name: 'Landscapes' }, removed: 3 }), 3)).toasts,
+        ['Removed 3 images from “Landscapes”.']);
+    // A failure keeps the listing; a deleted collection reloads it (the page then says the collection is gone).
+    const failed = await remove(() => { throw new TypeError('offline'); });
+    assert.deepEqual([failed.toasts, failed.changed], [['The images could not be removed. Try again.'], 0]);
+    const gone = await remove(() => { throw coded('collection-not-found', 404); });
+    assert.deepEqual([gone.toasts, gone.changed], [['That collection no longer exists.'], 1]);
+    const stale = await remove(() => { throw coded('source-changed', 409); });
+    assert.deepEqual([stale.toasts, stale.changed, stale.sourceChanged], [['The sources changed. Refresh the images and try again.'], 0, 1]);
+});

tests/js/workspace-images-state.test.js

@@ -211,11 +211,12 @@ test('custom collection pages reuse image-stream cancellation, cursors and ident
     assert.throws(()=>createImageStream({api:{get:async()=>{}},fetchPage:'invalid'}),/page/i);
 });
 
-test('stream sends category, tags and the rating sort and validates them', async () => {
+test('stream sends category, tags and the sort and validates them; the rating sort is gone (D153)', async () => {
     const {createImageStream}=await source;const calls=[];
     const stream=createImageStream({api:{get:async url=>{calls.push(new URL(url,'http://local').searchParams);return page([]);}}});
-    await stream.reset({q:'cat',category:'Animals',tags:['cats','studio'],sort:'rating'});
-    assert.equal(calls[0].get('category'),'Animals');assert.equal(calls[0].get('tags'),'cats,studio');assert.equal(calls[0].get('sort'),'rating');
+    await stream.reset({q:'cat',category:'Animals',tags:['cats','studio'],sort:'name-desc'});
+    assert.equal(calls[0].get('category'),'Animals');assert.equal(calls[0].get('tags'),'cats,studio');assert.equal(calls[0].get('sort'),'name-desc');
+    await assert.rejects(stream.reset({sort:'rating'}),/sort/i);
     await stream.reset({category:'',tags:[]});
     assert.equal(calls[1].has('category'),false);assert.equal(calls[1].has('tags'),false);
     for(const filters of [{category:3},{tags:'cats'},{tags:[1]},{tags:['a,b']},{tags:['']}]) await assert.rejects(stream.reset(filters),/filter/i);
@@ -277,7 +278,7 @@ test('the caption date names the newest result day under Newest first only', asy
     assert.equal(captionDate([at(2026,9,10),at(2026,9,16,23)],'oldest',now),'');
     assert.equal(captionDate([at(2026,9,13)],'newest',now),'Tuesday · October 13');
     assert.equal(captionDate([at(2025,2,3)],'newest',now),'Monday · March 3, 2025');
-    for(const sort of ['name','name-desc','rating']) assert.equal(captionDate([at(2026,9,17)],sort,now),'');
+    for(const sort of ['name','name-desc']) assert.equal(captionDate([at(2026,9,17)],sort,now),'');
     assert.equal(captionDate([],'newest',now),'');
     assert.equal(captionDate([{modified_time:'bad'}],'newest',now),'');
 });
@@ -525,7 +526,35 @@ test('defaults apply to a new visit but not to the current view or a saved view'
     assert.deepEqual(initialImagesView({ saved: { q: '' }, preferences }), { layout: 'grid', density: 'comfortable' });
     // Without readable preferences nothing changes.
     assert.deepEqual(initialImagesView({}), { layout: 'grid', density: 'comfortable' });
-    assert.deepEqual(initialImagesView({ preferences: {} }), { layout: 'grid', density: 'comfortable' });
+    // #451: a preference object without a layout falls back to the default layout, Masonry.
+    assert.deepEqual(initialImagesView({ preferences: {} }), { layout: 'masonry', density: 'comfortable' });
+});
+
+test('a default saved after the last visit started reaches the next visit unless the toolbar changed it (#451)', async () => {
+    const { initialImagesView } = await source;
+    const preferences = { workspaceLayout: 'masonry', workspaceDensity: 'compact' };
+    const started = { layout: 'grid', density: 'comfortable' };
+    // The visit kept the defaults it started from: the newer defaults win.
+    assert.deepEqual(initialImagesView({ prior: { layout: 'grid', density: 'comfortable', defaults: started }, preferences }),
+        { layout: 'masonry', density: 'compact' });
+    // A toolbar choice is the current view's own, whatever the defaults became; each value on its own.
+    assert.deepEqual(initialImagesView({ prior: { layout: 'list', density: 'comfortable', defaults: started }, preferences }),
+        { layout: 'list', density: 'compact' });
+    assert.deepEqual(initialImagesView({ prior: { layout: 'grid', density: 'large', defaults: started }, preferences }),
+        { layout: 'masonry', density: 'large' });
+    // Unchanged defaults: the visit returns as it was.
+    assert.deepEqual(initialImagesView({ prior: { layout: 'masonry', density: 'compact', defaults: { layout: 'masonry', density: 'compact' } }, preferences }),
+        { layout: 'masonry', density: 'compact' });
+    // A saved view and an embedded gallery never get a prior; without preferences the prior is kept.
+    assert.deepEqual(initialImagesView({ prior: { layout: 'grid', density: 'comfortable', defaults: started } }),
+        { layout: 'grid', density: 'comfortable' });
+});
+
+test('the defaults a visit starts from are the fresh-visit defaults (#451)', async () => {
+    const { visitDefaults } = await source;
+    assert.deepEqual(visitDefaults({ workspaceLayout: 'list', workspaceDensity: 'large' }), { layout: 'list', density: 'large' });
+    assert.deepEqual(visitDefaults({}), { layout: 'masonry', density: 'comfortable' });
+    assert.deepEqual(visitDefaults(null), { layout: 'grid', density: 'comfortable' });
 });
 
 test('listingFilter is the listing query as an object, without paging (#392)', async () => {

tests/js/workspace-library-editor-state.test.js

@@ -117,8 +117,9 @@ test('new prompt from Images waits for a fresh options epoch', async () => {
     assert.equal(sentBody.expected_epoch, EPOCH);
     assert.deepEqual(sentBody, {
         expected_epoch: EPOCH, positive_text: 'a portrait', negative_text: '',
-        parent_metadata: { title: 'Portrait', category: 'Uncategorized', tags: [], rating: null, notes: '' },
+        parent_metadata: { title: 'Portrait', category: 'Uncategorized', tags: [], notes: '' },
     });
+    assert.equal(Object.hasOwn(sentBody.parent_metadata, 'rating'), false, 'D153: a new prompt sends no rating');
     assert.equal(controller.state.phase, 'complete');
     assert.equal(committed.length, 1);
     assert.equal(committed[0].kind, 'create');
@@ -230,11 +231,11 @@ test('editor recovery preserves hyphenated ApiError codes', async () => {
         const server = harness(library({ write: () => response(409, { success: false, code, error: 'Conflict', library_epoch: EPOCH, ...extra }) }));
         const controller = createLibraryEditorController({ api: server.api, mode: 'metadata', positiveId: 7, pairId: 31 });
         await controller.initialize();
-        controller.setField('rating', 5);
+        controller.setField('notes', 'kept');
         await controller.submit();
         phases.push(controller.state.phase);
         if (controller.state.phase !== 'write-error') errors.push(controller.state.error);
-        assert.equal(controller.state.draft.rating, 5, 'draft preserved');
+        assert.equal(controller.state.draft.notes, 'kept', 'draft preserved');
         assert.equal(controller.state.receipt, null);
         controller.dispose();
     }
@@ -306,7 +307,7 @@ test('confirmed save refresh retry is read only', async () => {
     assert.equal(typeof createLibraryEditorController, 'function', 'controller export');
     let failReads = 1;
     const committed = [], reconciled = [];
-    const receipt = { success: true, library_epoch: EPOCH, positive: positive({ rating: null, notes: 'keep\ntext', edit_token: TOKEN2 }) };
+    const receipt = { success: true, library_epoch: EPOCH, positive: positive({ notes: 'keep\ntext', edit_token: TOKEN2 }) };
     const base = library({ write: () => response(200, receipt) });
     const server = harness(call => {
         if (call.method === 'GET' && server.writes().length && failReads-- > 0) return Promise.reject(new TypeError('offline'));
@@ -315,7 +316,6 @@ test('confirmed save refresh retry is read only', async () => {
     const controller = createLibraryEditorController({ api: server.api, mode: 'metadata', positiveId: 7, pairId: 31,
         onCommitted: commit => committed.push(commit), onReconciled: result => reconciled.push(result) });
     await controller.initialize();
-    controller.setField('rating', null);
     controller.setField('notes', 'keep\ntext');
     await controller.submit();
     assert.equal(controller.state.phase, 'saved-refresh-error');
@@ -338,7 +338,7 @@ test('confirmed save refresh retry is read only', async () => {
 test('legacy title fallback is not an edit', async () => {
     const { createLibraryEditorController } = await source;
     assert.equal(typeof createLibraryEditorController, 'function', 'controller export');
-    const receipt = { success: true, library_epoch: EPOCH, positive: positive({ rating: null, notes: 'keep\ntext' }) };
+    const receipt = { success: true, library_epoch: EPOCH, positive: positive({ notes: 'keep\ntext' }) };
     const server = harness(library({ write: () => response(200, receipt) }));
     const controller = createLibraryEditorController({ api: server.api, mode: 'metadata', positiveId: 7, pairId: 31 });
     await controller.initialize();
@@ -346,7 +346,6 @@ test('legacy title fallback is not an edit', async () => {
     assert.equal(controller.state.baseline.title, null, 'stored title stays distinct');
     assert.equal(controller.state.draft.tags, 'soft, warm');
     assert.equal(controller.state.draft.negative_text, 'blurry', 'the selected pair negative');
-    controller.setField('rating', null);
     controller.setField('notes', 'keep\ntext');
     controller.setField('tags', 'warm, soft,');
     controller.setField('category', ' Studio ');
@@ -357,7 +356,7 @@ test('legacy title fallback is not an edit', async () => {
     const sentBody = write.body;
     assert.equal(sentBody.expected_epoch, EPOCH);
     assert.equal(sentBody.if_match, TOKEN);
-    assert.deepEqual(sentBody.metadata, { rating: null, notes: 'keep\ntext' });
+    assert.deepEqual(sentBody.metadata, { notes: 'keep\ntext' });
 
     const unchanged = createLibraryEditorController({ api: server.api, mode: 'metadata', positiveId: 7, pairId: 31 });
     await unchanged.initialize();
@@ -421,12 +420,12 @@ test('metadata conflict reload keeps changed fields for review and never replays
     assert.equal(typeof createLibraryEditorController, 'function', 'controller export');
     let owner = positive();
     const server = harness(call => library({ owner, write: write => write.body.if_match === TOKEN2
-        ? response(200, { success: true, library_epoch: EPOCH, positive: { ...owner, rating: 5 } })
+        ? response(200, { success: true, library_epoch: EPOCH, positive: { ...owner, notes: 'mine' } })
         : response(409, { success: false, code: 'metadata-conflict', error: 'changed', library_epoch: EPOCH, positive: owner }) })(call));
     const controller = createLibraryEditorController({ api: server.api, mode: 'metadata', positiveId: 7, pairId: 31 });
     await controller.initialize();
-    controller.setField('rating', '5');
-    owner = positive({ title: 'Renamed elsewhere', notes: 'other writer', edit_token: TOKEN2 });
+    controller.setField('notes', 'mine');
+    owner = positive({ title: 'Renamed elsewhere', category: 'Elsewhere', edit_token: TOKEN2 });
     await controller.submit();
     assert.equal(controller.state.phase, 'conflict');
     await controller.submit();
@@ -434,20 +433,20 @@ test('metadata conflict reload keeps changed fields for review and never replays
     await controller.reloadOwner({ keepDraft: true });
     assert.equal(controller.state.phase, 'ready');
     assert.equal(controller.state.reviewRequired, true);
-    assert.equal(controller.state.draft.rating, 5, 'changed field kept');
+    assert.equal(controller.state.draft.notes, 'mine', 'changed field kept');
     assert.equal(controller.state.draft.title, 'Renamed elsewhere', 'untouched field follows the current owner');
-    assert.equal(controller.state.draft.notes, 'other writer');
+    assert.equal(controller.state.draft.category, 'Elsewhere');
     assert.equal(server.writes().length, 1);
     await controller.submit();
-    assert.deepEqual(server.writes()[1].body, { expected_epoch: EPOCH, if_match: TOKEN2, metadata: { rating: 5 } });
+    assert.deepEqual(server.writes()[1].body, { expected_epoch: EPOCH, if_match: TOKEN2, metadata: { notes: 'mine' } });
     assert.equal(controller.state.phase, 'complete');
     controller.dispose();
 
     const discard = createLibraryEditorController({ api: server.api, mode: 'metadata', positiveId: 7, pairId: 31 });
     await discard.initialize();
-    discard.setField('rating', 1);
+    discard.setField('notes', 'dropped');
     await discard.reloadOwner({ keepDraft: false });
-    assert.equal(discard.state.draft.rating, 3);
+    assert.equal(discard.state.draft.notes, '');
     assert.equal(discard.state.reviewRequired, false);
     discard.dispose();
 });
@@ -457,12 +456,12 @@ test('metadata unknown outcome retry reloads the owner instead of replaying', as
     assert.equal(typeof createLibraryEditorController, 'function', 'controller export');
     let owner = positive();
     const server = harness(call => library({ owner, write: () => {
-        owner = positive({ rating: 5, edit_token: TOKEN2 });
+        owner = positive({ notes: 'applied', edit_token: TOKEN2 });
         return Promise.reject(new TypeError('Failed to fetch'));
     } })(call));
     const controller = createLibraryEditorController({ api: server.api, mode: 'metadata', positiveId: 7, pairId: 31 });
     await controller.initialize();
-    controller.setField('rating', 5);
+    controller.setField('notes', 'applied');
     await controller.submit();
     assert.equal(controller.state.phase, 'unknown-outcome');
     await controller.retryUnconfirmed();
@@ -570,8 +569,10 @@ test('revision seeds text from the selected pair and sends Uncategorized for a b
     await controller.submit();
     assert.deepEqual(server.writes()[0].body, {
         expected_epoch: EPOCH, positive_text: 'First line\nrevised', negative_text: 'blurry',
-        parent_metadata: { title: 'Portrait', category: 'Uncategorized', tags: ['a', 'b'], rating: 3, notes: 'n' },
+        parent_metadata: { title: 'Portrait', category: 'Uncategorized', tags: ['a', 'b'], notes: 'n' },
     });
+    // D153: the revision's new positive is unrated; the owner's stored rating (3) is neither copied nor written.
+    assert.equal(Object.hasOwn(server.writes()[0].body.parent_metadata, 'rating'), false);
     controller.dispose();
 });
 
@@ -589,7 +590,9 @@ test('invalid drafts and server failures keep the draft without a receipt', asyn
     assert.equal(controller.state.phase, 'ready');
     assert.match(controller.state.error.message, /Title/);
     assert.throws(() => controller.setField('unknown', 'x'), TypeError);
-    assert.throws(() => controller.setField('rating', 7), TypeError);
+    // D153: the editor has no rating field.
+    assert.throws(() => controller.setField('rating', 4), /Unknown editor field: rating/);
+    assert.throws(() => controller.setField('notes', 4), TypeError);
     controller.setField('title', 'T');
     await controller.submit();
     assert.equal(controller.state.phase, 'write-error');
@@ -807,3 +810,44 @@ test('a blank category reused against an uncategorized owner is not unapplied me
         controller.dispose();
     }
 });
+
+// D153 (#447): the workspace no longer edits the 1–5 rating, and a save never clears a stored one.
+test('editing a prompt rated 4 sends no rating and the saved prompt keeps 4', async () => {
+    const { createLibraryEditorController, EDITOR_FIELDS } = await source;
+    assert.deepEqual(EDITOR_FIELDS, ['title', 'positive_text', 'negative_text', 'category', 'tags', 'notes']);
+    let owner = positive({ rating: 4, title: 'Portrait' });
+    const server = harness(call => library({ owner, write: write => {
+        owner = positive({ ...owner, ...write.body.metadata, edit_token: TOKEN2 });
+        return response(200, { success: true, library_epoch: EPOCH, positive: owner });
+    } })(call));
+    const reconciled = [];
+    const controller = createLibraryEditorController({ api: server.api, mode: 'metadata', positiveId: 7, pairId: 31,
+        onReconciled: result => reconciled.push(result) });
+    await controller.initialize();
+    assert.equal(Object.hasOwn(controller.state.draft, 'rating'), false, 'the draft has no rating');
+    assert.equal(Object.hasOwn(controller.state.baseline, 'rating'), false, 'the baseline has no rating');
+    controller.setField('title', 'Portrait, edited');
+    await controller.submit();
+    const [write] = server.writes();
+    assert.deepEqual(write.body.metadata, { title: 'Portrait, edited' });
+    assert.equal(controller.state.phase, 'complete');
+    assert.equal(controller.state.receipt.positive.rating, 4);
+    assert.equal(reconciled[0].detail.positive.rating, 4, 'the re-read prompt still has its rating');
+    controller.dispose();
+});
+
+test('a reused rated positive is not reported as unapplied metadata for its rating', async () => {
+    const { createLibraryEditorController } = await source;
+    const reused = recipe({ positive_created: false, pair_created: true, parent_metadata_applied: false,
+        positive: positive({ id: 9, text: 'a portrait', title: 'Portrait', category: 'Uncategorized', tags: [], rating: 4, notes: '' }) });
+    const server = harness(created(() => response(200, reused)));
+    const controller = createLibraryEditorController({ api: server.api, mode: 'new' });
+    await controller.initialize();
+    controller.setField('title', 'Portrait');
+    controller.setField('positive_text', 'a portrait');
+    await controller.submit();
+    assert.equal(Object.hasOwn(server.writes()[0].body.parent_metadata, 'rating'), false);
+    assert.equal(controller.state.phase, 'complete');
+    assert.equal(controller.state.outcome.metadataNotApplied, false);
+    controller.dispose();
+});

tests/js/workspace-library-format.test.js

@@ -46,9 +46,8 @@ test('used label boundaries in UTC and local time', async () => {
     for (const missing of [null, undefined, '', 'not a date', 42]) assert.equal(usedLabel(missing, new Date()), 'Not used yet');
 });
 
-test('rating label omits unrated', async () => {
-    const { ratingLabel } = await source;
-    assert.deepEqual(ratingLabel(5), { stars: '★★★★★', hidden: 'Rated 5 out of 5' });
-    assert.deepEqual(ratingLabel(1), { stars: '★', hidden: 'Rated 1 out of 5' });
-    for (const unrated of [null, undefined, 0, 6, -1, 2.5, '3', true]) assert.deepEqual(ratingLabel(unrated), { stars: '', hidden: '' });
+// D153 (#447): cards show no rating, so the layout module has no rating label.
+test('the layout module offers no rating label', async () => {
+    const layout = await source;
+    assert.equal(Object.hasOwn(layout, 'ratingLabel'), false);
 });

tests/js/workspace-library-selection.test.js

@@ -0,0 +1,96 @@
+const test = require('node:test');
+const assert = require('node:assert/strict');
+
+// #437 D148: the Prompt library's selection bar sets one category on the selected prompts, fenced by the storage epoch.
+const source = import('../../web/js/workspace/library-selection.mjs');
+
+test('the dialog is the shared simpleInput with the prompt title, and the route and cap are the legacy ones', async () => {
+    const {PROMPT_CATEGORY_DIALOG, CATEGORY_URL, MAX_CATEGORY_PROMPTS, TOO_MANY_PROMPTS} = await source;
+    assert.deepEqual({...PROMPT_CATEGORY_DIALOG}, {title:'Categorize selected prompts', label:'Category', action:'Apply'});
+    assert.equal(CATEGORY_URL, '/prompt_manager/bulk/category');
+    assert.equal(MAX_CATEGORY_PROMPTS, 1000);
+    assert.equal(TOO_MANY_PROMPTS, 'Select at most 1,000 prompts.');
+});
+
+test('a confirmed write toasts the shared outcome with "prompt", invalidates and reloads', async () => {
+    const {writeOutcome} = await source;
+    assert.deepEqual(writeOutcome({success:true, category:'Studio', updated:2, unchanged:1, not_found:0}, null),
+        {toast:'Category “Studio” set on 3 prompts.', committed:true, reload:true});
+    assert.deepEqual(writeOutcome({success:true, category:'Studio', updated:1, unchanged:0, not_found:2}, null),
+        {toast:'Category “Studio” set on 1 prompt. 2 could not be found.', committed:true, reload:true});
+});
+
+test('none found, a restored storage, an unconfirmed answer and a refused write each say what happened', async () => {
+    const {writeOutcome, NONE_FOUND, STORAGE_CHANGED, UNCONFIRMED, CATEGORY_FAILED} = await source;
+    assert.deepEqual(writeOutcome({success:true, category:'X', updated:0, unchanged:0, not_found:3}, null),
+        {toast:NONE_FOUND, committed:false, reload:true});
+    assert.deepEqual(writeOutcome(null, {status:409, code:'storage-changed'}), {toast:STORAGE_CHANGED, committed:false, reload:true});
+    for (const failure of [new TypeError('Failed to fetch'), {status:502}, {status:503}, {status:504}]) {
+        assert.deepEqual(writeOutcome(null, failure), {toast:UNCONFIRMED, committed:true, reload:true});
+    }
+    assert.deepEqual(writeOutcome(null, null), {toast:UNCONFIRMED, committed:true, reload:true});
+    for (const status of [400, 500]) {
+        assert.deepEqual(writeOutcome(null, {status}), {toast:CATEGORY_FAILED, committed:false, reload:false});
+    }
+    assert.equal(STORAGE_CHANGED, 'Prompt storage was restored since this page loaded. No category was changed.');
+});
+
+test('the field refuses control characters as validate_category does', async () => {
+    const {categoryProblem, CONTROL_CHARACTERS} = await source;
+    assert.equal(categoryProblem('Studio portraits'), null);
+    assert.equal(categoryProblem('tab\there'), CONTROL_CHARACTERS);
+    assert.equal(categoryProblem('nul\u0000'), CONTROL_CHARACTERS);
+    assert.equal(categoryProblem('unicode ✓ é'), null);
+});
+
+test('Add tags uses the Images dialog words for prompts, the tag routes and the editor limits (#442)', async () => {
+    const {PROMPT_TAGS_DIALOG, TAGS_URL, TAG_OPTIONS_URL, MAX_PROMPT_TAGS} = await source;
+    assert.deepEqual({...PROMPT_TAGS_DIALOG}, {title:'Tag selected prompts', label:'Tags, separated by commas', action:'Apply'});
+    assert.equal(TAGS_URL, '/prompt_manager/bulk/tags');
+    assert.equal(TAG_OPTIONS_URL, '/prompt_manager/tags');
+    assert.equal(MAX_PROMPT_TAGS, 20);
+});
+
+test('the Add tags toast says what was added, already present, skipped and missing', async () => {
+    const {tagsOutcome, writeOutcome, NONE_FOUND} = await source;
+    assert.equal(tagsOutcome({tags:['a', 'b'], updated:2, unchanged:0, over_limit:0, not_found:0}), 'Added “a, b” to 2 prompts.');
+    assert.equal(tagsOutcome({tags:['a'], updated:1, unchanged:2, over_limit:1, not_found:3}),
+        'Added “a” to 1 prompt. 2 prompts already had them. 1 prompt would pass 20 tags and was skipped. 3 could not be found.');
+    assert.equal(tagsOutcome({tags:['a'], updated:0, unchanged:2, over_limit:0, not_found:0}), '2 prompts already had “a”.');
+    // Every prompt skipped: found, so the skip sentence, never "none found"; nothing committed.
+    assert.deepEqual(writeOutcome({success:true, tags:['a'], updated:0, unchanged:0, over_limit:2, not_found:0}, null, 'tags'),
+        {toast:'2 prompts would pass 20 tags and were skipped.', committed:false, reload:true});
+    assert.deepEqual(writeOutcome({success:true, tags:['a'], updated:0, unchanged:0, over_limit:0, not_found:2}, null, 'tags'),
+        {toast:NONE_FOUND, committed:false, reload:true});
+});
+
+test('Add tags failures name tags, not the category', async () => {
+    const {writeOutcome, TAGS_STORAGE_CHANGED, TAGS_UNCONFIRMED, TAGS_FAILED} = await source;
+    assert.equal(writeOutcome(null, {status:409, code:'storage-changed'}, 'tags').toast, TAGS_STORAGE_CHANGED);
+    assert.equal(writeOutcome(null, {status:503}, 'tags').toast, TAGS_UNCONFIRMED);
+    assert.equal(writeOutcome(null, {status:400}, 'tags').toast, TAGS_FAILED);
+    assert.equal(TAGS_STORAGE_CHANGED, 'Prompt storage was restored since this page loaded. No tags were added.');
+});
+
+test('the Add tags field refuses what validate_tags refuses', async () => {
+    const {tagsProblem, TAG_PROBLEMS} = await source;
+    assert.equal(tagsProblem('red, blue'), null);
+    assert.equal(tagsProblem(' , ,'), TAG_PROBLEMS.none);
+    assert.equal(tagsProblem(Array.from({length:21}, (_, i) => `t${i}`).join(',')), TAG_PROBLEMS.many);
+    assert.equal(tagsProblem(Array.from({length:21}, () => 'same').join(',')), null);
+    assert.equal(tagsProblem('x'.repeat(51)), TAG_PROBLEMS.long);
+    assert.equal(tagsProblem('bad\ttag'), TAG_PROBLEMS.control);
+});
+
+test('Select all N loaded is offered while some loaded prompts are unchecked (#444 ruling)', async () => {
+    const {loadedOffer, offersLoaded} = await source;
+    assert.equal(loadedOffer(3), 'Select all 3 loaded');
+    assert.equal(loadedOffer(1203), 'Select all 1,203 loaded');
+    assert.equal(offersLoaded(new Set([1]), [1, 2, 3]), true);
+    assert.equal(offersLoaded(new Set([1, 2, 3]), [1, 2, 3]), false);
+    assert.equal(offersLoaded(new Set(), [1, 2]), false);
+    assert.equal(offersLoaded(new Set([9]), [9]), false);
+    // More loaded than any action accepts: never offered (Ctrl/Cmd+A still selects them).
+    assert.equal(offersLoaded(new Set([1]), Array.from({length:1001}, (_, i) => i + 1)), false);
+    assert.equal(offersLoaded(new Set([1]), Array.from({length:1000}, (_, i) => i + 1)), true);
+});

tests/js/workspace-library-state.test.js

@@ -5,8 +5,8 @@ const item=id=>({id,text:`Positive ${id}`,category:'Art',tags:['blue'],rating:3,
 const page=(items,cursor=null)=>({items,total:3,variant_total:6,has_more:cursor!==null,next_cursor:cursor});
 test('parent stream builds literal filters and deduplicates stable identities across infinite pages',async()=>{
  const {createLibraryStream}=await source;const calls=[];const pending=[page([item(1),item(2)],'next'),page([item(2),item(3)])];const s=createLibraryStream({api:{get:async(url)=>{calls.push(url);return pending.shift();}}});
- assert.equal(await s.reset({text:'a%_',category:'Art',tags:['blue','red'],sort:'rating'}),2);assert.equal(await s.loadMore(),1);assert.deepEqual(s.state.items.map(row=>row.id),[1,2,3]);assert.equal(s.state.hasMore,false);
- const query=new URL(calls[0],'https://x').searchParams;assert.equal(query.get('text'),'a%_');assert.deepEqual(query.getAll('tag'),['blue','red']);assert.equal(query.get('sort'),'rating');assert.equal(new URL(calls[1],'https://x').searchParams.get('cursor'),'next');assert.equal(await s.loadMore(),0);s.dispose();
+ assert.equal(await s.reset({text:'a%_',category:'Art',tags:['blue','red'],sort:'name'}),2);assert.equal(await s.loadMore(),1);assert.deepEqual(s.state.items.map(row=>row.id),[1,2,3]);assert.equal(s.state.hasMore,false);
+ const query=new URL(calls[0],'https://x').searchParams;assert.equal(query.get('text'),'a%_');assert.deepEqual(query.getAll('tag'),['blue','red']);assert.equal(query.get('sort'),'name');assert.equal(new URL(calls[1],'https://x').searchParams.get('cursor'),'next');assert.equal(await s.loadMore(),0);s.dispose();
 });
 test('the shelf caption total of negative variants comes from variant_total and resets with the filters',async()=>{
  const {createLibraryStream}=await source;const pending=[page([item(1)],'next'),{...page([item(2)]),variant_total:7},{...page([item(3)]),variant_total:1}];
@@ -104,3 +104,19 @@ test('archived variant reads use their own scope and cursor without changing the
  const active=[];const plain=createLibraryDetail({api:{get:async url=>{active.push(url);return {positive:item(1),variants:[]};}},onChange:()=>{}});await plain.load(1);assert.equal(new URL(active[0],'https://x').searchParams.has('archive'),false);
  assert.throws(()=>createLibraryDetail({api:{},onChange:()=>{},archive:'parents'}));
 });
+test('the stream keeps the first page\'s storage epoch and refuses a later page from another one (#437)',async()=>{
+ const {createLibraryStream}=await source;const A='11111111-1111-4111-8111-111111111111',B='22222222-2222-4222-8222-222222222222';
+ const pending=[{...page([item(1)],'next'),library_epoch:A},{...page([item(2)],'more'),library_epoch:A},{...page([item(3)]),library_epoch:B}];
+ const s=createLibraryStream({api:{get:async()=>pending.shift()}});assert.equal(s.state.epoch,null);
+ await s.reset();assert.equal(s.state.epoch,A);await s.loadMore();assert.equal(s.state.epoch,A);assert.equal(s.state.items.length,2);
+ await s.loadMore();assert.ok(s.state.error);assert.equal(s.state.items.length,2);assert.equal(s.state.epoch,A);
+ const restored=createLibraryStream({api:{},restore:{filters:{},state:s.state}});assert.equal(restored.state.epoch,A);restored.dispose();
+ const fresh=[{...page([item(4)]),library_epoch:B}];const reset=s.reset({text:'x'});assert.equal(s.state.epoch,null);
+ s.dispose();const next=createLibraryStream({api:{get:async()=>fresh.shift()}});await next.reset();assert.equal(next.state.epoch,B);next.dispose();await reset;
+});
+test('the shelf no longer sorts by rating (D153)',async()=>{
+ const {createLibraryStream}=await source;const s=createLibraryStream({api:{get:async()=>page([])}});
+ await assert.rejects(s.reset({sort:'rating'}),/Invalid library sort/);
+ assert.throws(()=>createLibraryStream({api:{get:async()=>page([])},restore:{filters:{sort:'rating'},state:{}}}),/Invalid library sort/);
+ for(const sort of ['newest','oldest','name']) assert.equal(await s.reset({sort}),0);s.dispose();
+});

tests/js/workspace-preferences.test.js

@@ -44,7 +44,7 @@ test('normalizes new workspace keys and keeps legacy keys', () => {
     // Invalid workspace values fall back to the defaults, as every legacy key does.
     store.text = JSON.stringify({ workspaceLayout: 'mosaic', workspaceDensity: 7, videoMute: 'no' });
     const fallback = module.readPreferences(() => store).values;
-    assert.equal(fallback.workspaceLayout, 'grid');
+    assert.equal(fallback.workspaceLayout, 'masonry'); // #451: the default layout
     assert.equal(fallback.workspaceDensity, 'comfortable');
     assert.equal(fallback.videoMute, true);
     // The legacy normalizer knows the new keys and still keeps unknown ones.
@@ -94,7 +94,7 @@ test('denied storage reports tab-only', () => {
     module.resetTabPreferences();
     const read = module.readPreferences(denied);
     assert.equal(read.persisted, false);
-    assert.equal(read.values.workspaceLayout, 'grid');
+    assert.equal(read.values.workspaceLayout, 'masonry');
     const result = module.writePreferences({ workspaceLayout: 'list', videoAutoplay: true }, denied);
     assert.deepEqual(result, { persisted: false, conflicts: [] });
     // The change is kept for this tab: every reader in it (a new Images visit, the viewer) sees it.

tests/js/workspace-saved-views.test.js

@@ -72,9 +72,9 @@ test('the listing keeps usable views in order and normalizes their state', async
 
 test('page filters and saved states convert both ways', async () => {
     const { imagesFilters, imagesViewState, libraryFilters, libraryViewState } = await source;
-    const filters = { q: 'castle', category: 'Portraits', sort: 'rating', root: '1', subfolder: 'x', media: 'image', tags: ['fog'], favorites: true, collection: '' };
+    const filters = { q: 'castle', category: 'Portraits', sort: 'name-desc', root: '1', subfolder: 'x', media: 'image', tags: ['fog'], favorites: true, collection: '' };
     const state = imagesViewState(filters);
-    assert.deepEqual(state, { q: 'castle', category: 'Portraits', sort: 'rating', source: '1', folder: 'x', media: 'image', tags: ['fog'], favorites: true, collection: '' });
+    assert.deepEqual(state, { q: 'castle', category: 'Portraits', sort: 'name-desc', source: '1', folder: 'x', media: 'image', tags: ['fog'], favorites: true, collection: '' });
     assert.deepEqual(imagesFilters(state), filters);
     assert.deepEqual(imagesFilters({}), { q: '', category: '', sort: 'newest', root: 'all', subfolder: '', media: 'all', tags: [], favorites: false, collection: '' });
     // #331 D140: the open collection round-trips; a missing one is ''.
@@ -215,3 +215,13 @@ test('a state over the API 4 KB cap is refused before anything is sent (#249 M5)
     await assert.rejects(store.create('images', 'Big', big), error => error instanceof SavedViewTooLarge && error instanceof RangeError);
     assert.deepEqual(api.calls, []);
 });
+
+test('a view saved under the retired Highest rated sort opens on Recently added (D153)', async () => {
+    const { normalizeViewState, imagesFilters, libraryFilters } = await source;
+    assert.equal(normalizeViewState('images', { sort: 'rating' }).sort, 'newest');
+    assert.equal(normalizeViewState('library', { sort: 'rating' }).sort, 'newest');
+    assert.equal(imagesFilters({ sort: 'rating' }).sort, 'newest');
+    assert.equal(libraryFilters({ sort: 'rating' }).sort, 'newest');
+    assert.equal(normalizeViewState('images', { sort: 'name-desc' }).sort, 'name-desc');
+    assert.equal(normalizeViewState('library', { sort: 'oldest' }).sort, 'oldest');
+});

tests/js/workspace-system-appearance.test.js

@@ -162,7 +162,8 @@ test('reset restores defaults without touching other keys', async () => {
         const reset = dom.find(surface.children.at(-1), node => node.tagName === 'BUTTON');
         assert.equal(reset.textContent, 'Reset');
         reset.click();
-        assert.deepEqual(drafts.changes('browser'), { workspaceLayout: 'grid', videoMute: true, showImageInfo: true, sidebarCollapsedByDefault: false });
+        // #451: the default layout is Masonry.
+        assert.deepEqual(drafts.changes('browser'), { workspaceLayout: 'masonry', videoMute: true, showImageInfo: true, sidebarCollapsedByDefault: false });
         // The host fields and every key outside the display defaults are left as they are.
         assert.deepEqual(drafts.changes('host'), { result_timeout: 9 });
         assert.deepEqual([...DISPLAY_KEYS].sort(), Object.keys(BROWSER).sort());

tests/js/workspace-system-backups.test.js

@@ -192,6 +192,71 @@ test('views reload once after a restore this page saw finish; a failed restore c
     assert.equal(listeners.size, 0);
 });
 
+// #415: a fake document whose open dialogs the test sets, and a MutationObserver stand-in it triggers.
+function dialogPage() {
+    const page = { open: 0, observers: new Set() };
+    page.doc = { documentElement: {}, querySelector: selector => (selector === 'dialog[open]' && page.open > 0 ? {} : null) };
+    page.Observer = class {
+        constructor(callback) { this.callback = callback; }
+        observe(target, options) { this.target = target; this.options = options; page.observers.add(this); }
+        disconnect() { page.observers.delete(this); }
+    };
+    page.mutate = open => { page.open = open; for (const observer of [...page.observers]) observer.callback([]); };
+    return page;
+}
+
+test('the restore reload runs at once when no dialog is open', async () => {
+    const { createRestoreReload } = await backups;
+    const page = dialogPage();
+    let reloads = 0;
+    const reload = createRestoreReload(() => { reloads += 1; }, { doc: page.doc, Observer: page.Observer });
+    reload.request();
+    assert.equal(reloads, 1);
+    assert.equal(page.observers.size, 0, 'nothing waits');
+    reload.request();
+    assert.equal(reloads, 2);
+});
+
+test('the restore reload waits for every open dialog to close, then runs once (#415)', async () => {
+    const { createRestoreReload } = await backups;
+    const page = dialogPage();
+    page.open = 1;
+    let reloads = 0;
+    const reload = createRestoreReload(() => { reloads += 1; }, { doc: page.doc, Observer: page.Observer });
+    reload.request();
+    reload.request();
+    assert.equal(reloads, 0, 'an open dialog holds the reload');
+    assert.equal(page.observers.size, 1, 'one observer for repeated requests');
+    const [observer] = page.observers;
+    assert.equal(observer.target, page.doc.documentElement);
+    assert.deepEqual(observer.options, { subtree: true, childList: true, attributes: true, attributeFilter: ['open'] });
+    page.mutate(1);
+    assert.equal(reloads, 0, 'a dialog swapped for another keeps it waiting');
+    page.mutate(0);
+    assert.equal(reloads, 1);
+    assert.equal(page.observers.size, 0, 'the observer stops once it ran');
+    page.mutate(0);
+    assert.equal(reloads, 1);
+    page.open = 1;
+    reload.request();
+    page.mutate(0);
+    assert.equal(reloads, 2, 'a later restore waits again');
+});
+
+test('disposing drops a waiting restore reload', async () => {
+    const { createRestoreReload } = await backups;
+    const page = dialogPage();
+    page.open = 1;
+    let reloads = 0;
+    const reload = createRestoreReload(() => { reloads += 1; }, { doc: page.doc, Observer: page.Observer });
+    reload.request();
+    reload.dispose();
+    assert.equal(page.observers.size, 0);
+    page.mutate(0);
+    assert.equal(reloads, 0);
+    reload.dispose();
+});
+
 test('a downloaded backup the server could not verify says so', async () => {
     const { downloadDatabaseBackup, BACKUP_DONE, backupUnverified } = await scan;
     const toasts = [];

tests/js/workspace-system-overview.test.js

@@ -141,7 +141,7 @@ test('the report summary says when it was checked and whether it is complete', a
     assert.equal(healthSummary({ ...REPORT, checked_at: 'x' }), 'Nothing was changed; review the checks with issues.');
 });
 
-test('statistics rows are the real runs and ratings, then the library totals', async () => {
+test('statistics rows are the real runs, then the library totals, without ratings (D153)', async () => {
     const { categoryRows, totalsRows } = await overviewModule;
     const statistics = {
         by_category: [{ category: 'Landscapes', runs: 1200, average_rating: 4.5 },
@@ -149,8 +149,8 @@ test('statistics rows are the real runs and ratings, then the library totals', a
         totals: { total_prompts: 12, total_tags: 15, total_images: 24 },
         rated_images: 20,
     };
-    assert.deepEqual(categoryRows(statistics), [['Landscapes', (1200).toLocaleString(), '4.5 / 5'], ['Uncategorized', '3', 'Not rated']]);
-    assert.deepEqual(totalsRows(statistics), [['Prompts', '12'], ['Tags', '15'], ['Images', '24'], ['Rated images', '20']]);
+    assert.deepEqual(categoryRows(statistics), [['Landscapes', (1200).toLocaleString()], ['Uncategorized', '3']]);
+    assert.deepEqual(totalsRows(statistics), [['Prompts', '12'], ['Tags', '15'], ['Images', '24']]);
 });
 
 /** A fake api answering each `get`/`post` from a queue per URL, recording the calls. */

tests/js/workspace-toolbar.test.js

@@ -594,6 +594,9 @@ test('an input dialog offers suggestions and a secondary action', async () => {
         const [plainField, plainActions] = byId(dom, 'modal-body').children;
         assert.equal(plainField.children[1].getAttribute('list'), null);
         assert.deepEqual(plainActions.children.map(node => node.textContent), ['Cancel', 'A']);
+        // An empty list (Add to collection with no collections yet, #331 D138) adds no datalist either.
+        openInputDialog({ scope, title: 'T', label: 'L', action: 'A', suggestions: [], onApply() {} });
+        assert.equal(dom.all(node => node.tagName === 'DATALIST').length, 0);
         assert.throws(() => openInputDialog({ scope, title: 'T', label: 'L', action: 'A', suggestions: [1], onApply() {} }), TypeError);
         assert.throws(() => openInputDialog({ scope, title: 'T', label: 'L', action: 'A', secondary: { label: '' }, onApply() {} }), TypeError);
         assert.throws(() => openInputDialog({ scope, title: 'T', label: 'L', action: 'A', secondary: { label: 'X' }, onApply() {} }), TypeError);
@@ -908,6 +911,34 @@ test('openReviewDialog lays out heading, grid, options, sentence, backup box and
     }
 });
 
+// #415: a finished restore ends an open review (its reviewed plan no longer describes storage) as a dismissal, and
+// leaves any other dialog, with what was typed in it, open.
+test('closeReviewDialog closes only an open review, as dismissed', async () => {
+    const { openReviewDialog, openInputDialog, closeReviewDialog, closeWorkspaceDialog } = await dialogsModule;
+    const { createScope } = await coreModule;
+    const dom = fakeDOM();
+    const scope = createScope();
+    try {
+        closeReviewDialog();
+        const outcomes = [];
+        openInputDialog({ scope, title: 'Save view', label: 'Name', value: 'typed', action: 'Save', onApply() {} });
+        closeReviewDialog();
+        assert.equal(modal(dom).open, true, 'an input dialog stays');
+        assert.equal(byId(dom, 'action-input').value, 'typed');
+        openReviewDialog({ scope, title: 'Restore preview', grid: [['Integrity', 'Verified']], cancelLabel: 'Cancel',
+            primary: { label: 'Restore', danger: true }, onStart: () => outcomes.push('start'), onDismiss: () => outcomes.push('dismiss') });
+        closeReviewDialog();
+        assert.deepEqual(outcomes, ['dismiss']);
+        assert.equal(modal(dom), undefined, 'the shell is gone');
+        closeReviewDialog();
+        assert.deepEqual(outcomes, ['dismiss']);
+    } finally {
+        scope.dispose();
+        closeWorkspaceDialog();
+        dom.restore();
+    }
+});
+
 test('review dialog backup box keeps the dialog open and toasts the outcome', async () => {
     const { openReviewDialog, closeWorkspaceDialog, SCAN_BACKUP_COPY } = await dialogsModule;
     const { downloadDatabaseBackup, BACKUP_DONE, BACKUP_FAILED } = await scanModule;

tests/test_browser_fast_reset.py

@@ -22,6 +22,9 @@ from database.operations import PromptDatabase
 from tests.browser import fast_reset
 from tests.browser.fast_reset import ScratchSeeder, table_digests
 
+# The owner lock of /dev/shm scratch folders needs POSIX flock (#419).
+HAS_FLOCK = unittest.skipUnless(os.name == "posix", "needs POSIX flock")
+
 
 def digests(db: PromptDatabase) -> dict:
     return table_digests(db.model.get_connection())
@@ -275,6 +278,7 @@ class TestScratchSeeder(unittest.TestCase):
         self.assertEqual(digests(self.live), expected)
         self.assertFalse((self.root / "unused").exists())
 
+    @HAS_FLOCK
     def test_the_scratch_lives_under_the_memory_root_when_there_is_one(self):
         memory = self.root / "memory"
         memory.mkdir()
@@ -288,6 +292,16 @@ class TestScratchSeeder(unittest.TestCase):
         seeder.close()
         self.assertEqual(list(memory.iterdir()), [])
 
+    @unittest.skipIf(os.name == "posix", "without flock the memory root is unused")
+    def test_without_flock_the_scratch_stays_on_disk(self):
+        memory = self.root / "memory"
+        memory.mkdir()
+        seeder = ScratchSeeder(self.api, self.root / "fallback", memory_root=memory)
+        self.addCleanup(seeder.close)
+        seeder.run(reseed)
+        self.assertEqual(list(memory.iterdir()), [])
+        self.assertTrue((self.root / "fallback" / "seed.db").exists())
+
     def test_a_full_memory_root_puts_the_scratch_on_disk(self):
         memory = self.root / "memory"
         memory.mkdir()
@@ -303,26 +317,32 @@ class TestScratchSeeder(unittest.TestCase):
         self.assertEqual(list(memory.iterdir()), [])
         self.assertTrue((self.root / "fallback" / "seed.db").exists())
 
-    @unittest.skipUnless(os.name == "posix", "process probe is POSIX only")
-    def test_scratch_folders_of_dead_servers_are_removed(self):
+    @HAS_FLOCK
+    def test_the_scratch_folder_is_held_by_a_lock_until_close(self):
         memory = self.root / "memory"
         memory.mkdir()
-        dead = subprocess.Popen([sys.executable, "-c", "pass"])
-        dead.wait()
-        orphan = memory / f"pm-reset-{dead.pid}-abc"
-        alive = memory / f"pm-reset-{os.getpid()}-def"
-        other = memory / "pm-reset-notapid"
-        for folder in (orphan, alive, other):
-            folder.mkdir()
-            (folder / "seed.db").write_bytes(b"x")
         seeder = ScratchSeeder(self.api, self.root / "fallback", memory_root=memory)
         self.addCleanup(seeder.close)
         seeder.run(reseed)
-        self.assertFalse(orphan.exists())
-        self.assertTrue(alive.exists())
-        self.assertTrue(other.exists())
-        own = [p for p in memory.glob(f"pm-reset-{os.getpid()}-*") if p != alive]
-        self.assertEqual(len(own), 1)
+        (folder,) = memory.iterdir()
+        self.assertTrue(folder.name.startswith(f"pm-reset-held-{os.getpid()}-"))
+        self.assertFalse(lock_is_free(folder / "owner.lock"))
+        seeder.close()
+        self.assertEqual(list(memory.iterdir()), [])
+
+    @HAS_FLOCK
+    def test_a_lock_that_cannot_be_taken_puts_the_scratch_on_disk(self):
+        memory = self.root / "memory"
+        memory.mkdir()
+        seeder = ScratchSeeder(self.api, self.root / "fallback", memory_root=memory)
+        self.addCleanup(seeder.close)
+        refuse = mock.patch.object(
+            fast_reset.fcntl, "flock", side_effect=OSError("no locks here")
+        )
+        with refuse:
+            seeder.run(reseed)
+        self.assertEqual(list(memory.iterdir()), [])
+        self.assertTrue((self.root / "fallback" / "seed.db").exists())
 
     def test_close_releases_the_scratch_files(self):
         self.seeder.run(reseed)
@@ -331,5 +351,95 @@ class TestScratchSeeder(unittest.TestCase):
         self.seeder.close()  # idempotent
 
 
+# A process that takes the owner lock of a folder, says so, and waits.
+HOLDER = """
+import fcntl, os, sys, time
+fd = os.open(os.path.join(sys.argv[1], "owner.lock"), os.O_RDWR | os.O_CREAT)
+fcntl.flock(fd, fcntl.LOCK_EX)
+print("held", flush=True)
+time.sleep(600)
+"""
+
+
+def lock_is_free(path: Path) -> bool:
+    fd = os.open(path, os.O_RDWR)
+    try:
+        fast_reset.fcntl.flock(fd, fast_reset.fcntl.LOCK_EX | fast_reset.fcntl.LOCK_NB)
+    except BlockingIOError:
+        return False
+    finally:
+        os.close(fd)
+    return True
+
+
+@HAS_FLOCK
+class TestOrphanedScratch(unittest.TestCase):
+    """Only folders whose owner lock is free are orphans (#419)."""
+
+    def setUp(self):
+        self.tmp = tempfile.TemporaryDirectory()
+        self.addCleanup(self.tmp.cleanup)
+        self.memory = Path(self.tmp.name)
+
+    def folder(self, name: str) -> Path:
+        folder = self.memory / name
+        folder.mkdir()
+        (folder / "seed.db").write_bytes(b"x")
+        return folder
+
+    def hold(self, folder: Path) -> subprocess.Popen:
+        holder = subprocess.Popen(
+            [sys.executable, "-c", HOLDER, str(folder)],
+            stdout=subprocess.PIPE,
+            text=True,
+        )
+        self.addCleanup(holder.wait)
+        self.addCleanup(holder.stdout.close)
+        self.addCleanup(holder.kill)
+        self.assertEqual(holder.stdout.readline().strip(), "held")
+        return holder
+
+    def test_a_folder_whose_owner_was_killed_is_removed(self):
+        folder = self.folder("pm-reset-held-123-abc")
+        holder = self.hold(folder)
+        holder.kill()
+        holder.wait()
+        fast_reset.remove_orphaned_scratch(self.memory)
+        self.assertFalse(folder.exists())
+
+    def test_a_held_folder_is_kept_whatever_pid_its_name_carries(self):
+        # Under --ipc=host the owner's pid is from another PID namespace: it
+        # does not exist here, but its lock does.
+        dead = subprocess.Popen([sys.executable, "-c", "pass"])
+        dead.wait()
+        folder = self.folder(f"pm-reset-held-{dead.pid}-abc")
+        self.hold(folder)
+        fast_reset.remove_orphaned_scratch(self.memory)
+        self.assertTrue(folder.exists())
+        self.assertTrue((folder / "seed.db").exists())
+
+    def test_folders_without_an_owner_lock_are_kept(self):
+        dead = subprocess.Popen([sys.executable, "-c", "pass"])
+        dead.wait()
+        legacy = self.folder(f"pm-reset-{dead.pid}-abc")
+        staging = self.folder(f"pm-reset-new-{dead.pid}-def")
+        unlocked = self.folder(f"pm-reset-held-{dead.pid}-ghi")
+        fast_reset.remove_orphaned_scratch(self.memory)
+        for folder in (legacy, staging, unlocked):
+            self.assertTrue(folder.exists(), folder.name)
+
+    def test_a_folder_held_by_another_seeder_in_this_process_is_kept(self):
+        api = types.SimpleNamespace(db=PromptDatabase(str(self.memory / "live.db")))
+        self.addCleanup(api.db.close_all)
+        root = self.memory / "shm"
+        root.mkdir()
+        other = ScratchSeeder(api, self.memory / "fallback", memory_root=root)
+        self.addCleanup(other.close)
+        other.run(reseed)
+        (folder,) = root.iterdir()
+        fast_reset.remove_orphaned_scratch(root)
+        self.assertTrue(folder.exists())
+
+
 if __name__ == "__main__":
     unittest.main()

tests/test_bulk_library_category.py

@@ -0,0 +1,255 @@
+"""#437: the workspace Library's fenced bulk "Set category" (DB command and route)."""
+
+import sqlite3
+import tempfile
+import unittest
+from pathlib import Path
+
+from database.execution_history_schema import rotate_history_epoch
+from database.library_commands import LibraryCommandError
+from database.operations import PromptDatabase
+from utils.hashing import generate_prompt_hash
+
+try:
+    from test_api_prompts import RouteCoverageCase
+except ImportError:
+    from tests.test_api_prompts import RouteCoverageCase
+
+URL = "/prompt_manager/bulk/category"
+
+
+def save(db, text, **metadata):
+    return db.save_prompt(text, prompt_hash=generate_prompt_hash(text), **metadata)
+
+
+def row(db, prompt_id):
+    with db.model.get_connection() as conn:
+        found = conn.execute(
+            "SELECT category, updated_at FROM prompts WHERE id=?", (prompt_id,)
+        ).fetchone()
+    return tuple(found) if found else None
+
+
+class TestBulkLibraryCategory(unittest.TestCase):
+    def setUp(self):
+        self.temp = tempfile.TemporaryDirectory()
+        self.db = PromptDatabase(str(Path(self.temp.name) / "bulk.db"))
+        self.epoch = self.db.get_library_options()["library_epoch"]
+
+    def tearDown(self):
+        self.db.close_all()
+        self.temp.cleanup()
+
+    def run_bulk(self, ids, category, epoch=None):
+        return self.db.bulk_set_library_category(
+            ids, category, expected_epoch=epoch or self.epoch
+        )
+
+    def test_sets_changed_rows_and_counts_unchanged_and_missing(self):
+        a = save(self.db, "Alpha", category="Old")
+        b = save(self.db, "Beta", category="New")
+        c = save(self.db, "Gamma")
+        before_b = row(self.db, b)
+        result = self.run_bulk([a, b, c, 99999], "  New ")
+        self.assertEqual(
+            result,
+            {
+                "library_epoch": self.epoch,
+                "category": "New",
+                "updated": 2,
+                "unchanged": 1,
+                "not_found": 1,
+            },
+        )
+        self.assertEqual(row(self.db, a)[0], "New")
+        self.assertEqual(row(self.db, c)[0], "New")
+        # An unchanged row keeps its timestamp (and so its edit token).
+        self.assertEqual(row(self.db, b), before_b)
+
+    def test_changed_category_rotates_the_edit_token_only_for_changed_rows(self):
+        a = save(self.db, "Alpha", category="Old")
+        b = save(self.db, "Beta", category="New")
+        tokens = {
+            pid: self.db.get_library_positive(pid)["positive"]["edit_token"]
+            for pid in (a, b)
+        }
+        self.run_bulk([a, b], "New")
+        self.assertNotEqual(
+            self.db.get_library_positive(a)["positive"]["edit_token"], tokens[a]
+        )
+        self.assertEqual(
+            self.db.get_library_positive(b)["positive"]["edit_token"], tokens[b]
+        )
+
+    def test_changed_rows_become_registered_library_parents_like_the_editor(self):
+        a = save(self.db, "Alpha", category="Old")
+        b = save(self.db, "Beta", category="Same")
+
+        def registered(pid):
+            with self.db.model.get_connection() as conn:
+                return bool(
+                    conn.execute(
+                        "SELECT 1 FROM prompt_library_state WHERE prompt_id=?", (pid,)
+                    ).fetchone()
+                )
+
+        self.assertEqual((registered(a), registered(b)), (False, False))
+        self.run_bulk([a, b], "Same")
+        self.assertEqual((registered(a), registered(b)), (True, False))
+
+    def test_updated_at_is_utc_iso_text(self):
+        a = save(self.db, "Alpha")
+        self.run_bulk([a], "Utc")
+        self.assertTrue(row(self.db, a)[1].endswith("+00:00"))
+
+    def test_negative_only_prompts_are_not_library_parents(self):
+        pair = self.db.save_prompt_pair("Cat", "Blur")
+        negative = self.db.get_prompt_pair(pair)["negative_prompt_id"]
+        positive = self.db.get_prompt_pair(pair)["positive_prompt_id"]
+        result = self.run_bulk([positive, negative], "Animals")
+        self.assertEqual((result["updated"], result["not_found"]), (1, 1))
+        self.assertIsNone(row(self.db, negative)[0])
+
+    def test_restored_storage_writes_nothing(self):
+        a = save(self.db, "Alpha", category="Old")
+        new_epoch = rotate_history_epoch(self.db.model.get_connection())
+        with self.assertRaises(LibraryCommandError) as caught:
+            self.run_bulk([a], "New")
+        self.assertEqual(caught.exception.code, "storage-changed")
+        self.assertEqual(caught.exception.epoch, new_epoch)
+        self.assertEqual(row(self.db, a)[0], "Old")
+
+    def test_a_failure_mid_write_leaves_every_row_as_it_was(self):
+        ids = [save(self.db, f"P{i}", category="Old") for i in range(3)]
+        with self.db.model.get_connection() as conn:
+            conn.execute(
+                "CREATE TRIGGER fail_third BEFORE UPDATE OF category ON prompts "
+                f"WHEN OLD.id={ids[2]} BEGIN SELECT RAISE(ABORT, 'boom'); END"
+            )
+        with self.assertRaises(sqlite3.DatabaseError):
+            self.run_bulk(ids, "New")
+        self.assertEqual([row(self.db, pid)[0] for pid in ids], ["Old"] * 3)
+
+    def test_many_ids_are_read_in_chunks(self):
+        ids = [save(self.db, f"Many {i}") for i in range(1000)]
+        result = self.run_bulk(ids, "Bulk")
+        self.assertEqual(result["updated"], 1000)
+
+    def test_rejects_invalid_arguments(self):
+        a = save(self.db, "Alpha")
+        for ids, category, epoch in (
+            ([a], "", self.epoch),
+            ([a], "   ", self.epoch),
+            ([a], None, self.epoch),
+            ([a], "x" * 101, self.epoch),
+            ([a], "bad\x01", self.epoch),
+            ([a], "ok", "not-a-uuid"),
+            ([], "ok", self.epoch),
+        ):
+            with self.subTest(category=category, epoch=epoch, ids=ids):
+                with self.assertRaises(ValueError):
+                    self.run_bulk(ids, category, epoch)
+
+
+class TestBulkLibraryCategoryRoute(RouteCoverageCase):
+    def epoch(self):
+        return self.api.db.get_library_options()["library_epoch"]
+
+    async def test_fenced_form_sets_categories_and_reports_counts(self):
+        ids = [self._save_prompt(f"route {i}", category="Old") for i in range(2)]
+        status, body = await self._json(
+            "POST",
+            URL,
+            json={
+                "prompt_ids": ids,
+                "category": "Fresh",
+                "expected_epoch": self.epoch(),
+            },
+        )
+        self.assertEqual(status, 200)
+        self.assertEqual(
+            body,
+            {
+                "success": True,
+                "library_epoch": self.epoch(),
+                "category": "Fresh",
+                "updated": 2,
+                "unchanged": 0,
+                "not_found": 0,
+            },
+        )
+        self.assertEqual(self.api.db.get_prompt_by_id(ids[1])["category"], "Fresh")
+
+    async def test_restored_storage_answers_409_storage_changed(self):
+        pid = self._save_prompt("route restore", category="Old")
+        stale = self.epoch()
+        new_epoch = rotate_history_epoch(self.api.db.model.get_connection())
+        status, body = await self._json(
+            "POST",
+            URL,
+            json={"prompt_ids": [pid], "category": "Fresh", "expected_epoch": stale},
+        )
+        self.assertEqual((status, body["code"]), (409, "storage-changed"))
+        self.assertFalse(body["success"])
+        self.assertEqual(body["library_epoch"], new_epoch)
+        self.assertIsInstance(body["error"], str)
+        self.assertEqual(self.api.db.get_prompt_by_id(pid)["category"], "Old")
+
+    async def test_fenced_form_validates_its_body(self):
+        pid = self._save_prompt("route validate")
+        epoch = self.epoch()
+        for body in (
+            {"prompt_ids": [pid], "category": "", "expected_epoch": epoch},
+            {"prompt_ids": [pid], "category": "  ", "expected_epoch": epoch},
+            {"prompt_ids": [pid], "category": 5, "expected_epoch": epoch},
+            {"prompt_ids": [pid], "category": "x" * 101, "expected_epoch": epoch},
+            {"prompt_ids": [pid], "category": "ok", "expected_epoch": "nope"},
+            {"prompt_ids": [pid], "category": "ok", "expected_epoch": 7},
+            {"prompt_ids": [], "category": "ok", "expected_epoch": epoch},
+            {"prompt_ids": [True], "category": "ok", "expected_epoch": epoch},
+            {
+                "prompt_ids": list(range(1, 1002)),
+                "category": "ok",
+                "expected_epoch": epoch,
+            },
+            {"prompt_ids": [pid], "expected_epoch": epoch},
+            {
+                "prompt_ids": [pid],
+                "category": "ok",
+                "expected_epoch": epoch,
+                "extra": 1,
+            },
+        ):
+            with self.subTest(body=body):
+                status, answer = await self._json("POST", URL, json=body)
+                self.assertEqual(status, 400)
+                self.assertFalse(answer["success"])
+        self.assertIsNone(self.api.db.get_prompt_by_id(pid)["category"])
+
+    async def test_fenced_form_reports_an_unexpected_failure(self):
+        pid = self._save_prompt("route failure")
+        self._break("bulk_set_library_category")
+        status, body = await self._json(
+            "POST",
+            URL,
+            json={
+                "prompt_ids": [pid],
+                "category": "ok",
+                "expected_epoch": self.epoch(),
+            },
+        )
+        self.assertEqual(status, 500)
+        self.assertFalse(body["success"])
+
+    async def test_legacy_form_is_unchanged(self):
+        pid = self._save_prompt("route legacy", category="Old")
+        status, body = await self._json(
+            "POST", URL, json={"prompt_ids": [pid], "category": "Legacy"}
+        )
+        self.assertEqual((status, body["updated_count"]), (200, 1))
+        self.assertNotIn("library_epoch", body)
+        self.assertEqual(self.api.db.get_prompt_by_id(pid)["category"], "Legacy")
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_bulk_library_tags.py

@@ -0,0 +1,286 @@
+"""#442: the workspace Library's fenced bulk "Add tags" (DB command and route)."""
+
+import sqlite3
+import tempfile
+import unittest
+from pathlib import Path
+
+from database.execution_history_schema import rotate_history_epoch
+from database.library_commands import LibraryCommandError
+from database.operations import PromptDatabase
+from utils.hashing import generate_prompt_hash
+
+try:
+    from test_api_prompts import RouteCoverageCase
+except ImportError:
+    from tests.test_api_prompts import RouteCoverageCase
+
+URL = "/prompt_manager/bulk/tags"
+
+
+def save(db, text, **metadata):
+    return db.save_prompt(text, prompt_hash=generate_prompt_hash(text), **metadata)
+
+
+def tags_of(db, prompt_id):
+    with db.model.get_connection() as conn:
+        rows = conn.execute(
+            "SELECT t.name FROM prompt_tags pt JOIN tags t ON t.id=pt.tag_id "
+            "WHERE pt.prompt_id=? ORDER BY t.name",
+            (prompt_id,),
+        ).fetchall()
+    return [row[0] for row in rows]
+
+
+def updated_at(db, prompt_id):
+    with db.model.get_connection() as conn:
+        return conn.execute(
+            "SELECT updated_at FROM prompts WHERE id=?", (prompt_id,)
+        ).fetchone()[0]
+
+
+def tag_names(db):
+    with db.model.get_connection() as conn:
+        return sorted(row[0] for row in conn.execute("SELECT name FROM tags"))
+
+
+class TestBulkLibraryTags(unittest.TestCase):
+    def setUp(self):
+        self.temp = tempfile.TemporaryDirectory()
+        self.db = PromptDatabase(str(Path(self.temp.name) / "tags.db"))
+        self.epoch = self.db.get_library_options()["library_epoch"]
+
+    def tearDown(self):
+        self.db.close_all()
+        self.temp.cleanup()
+
+    def run_bulk(self, ids, tags, epoch=None):
+        return self.db.bulk_add_library_tags(
+            ids, tags, expected_epoch=epoch or self.epoch
+        )
+
+    def test_adds_only_missing_tags_and_counts_each_outcome(self):
+        a = save(self.db, "Alpha", tags=["red"])
+        b = save(self.db, "Beta", tags=["red", "blue"])
+        c = save(self.db, "Gamma")
+        before_b = updated_at(self.db, b)
+        result = self.run_bulk([a, b, c, 99999], [" red ", "blue", "Blue"])
+        self.assertEqual(
+            result,
+            {
+                "library_epoch": self.epoch,
+                "tags": ["red", "blue"],
+                "updated": 2,
+                "unchanged": 1,
+                "not_found": 1,
+                "over_limit": 0,
+            },
+        )
+        self.assertEqual(tags_of(self.db, a), ["blue", "red"])
+        self.assertEqual(tags_of(self.db, c), ["blue", "red"])
+        self.assertEqual(updated_at(self.db, b), before_b)
+        self.assertTrue(updated_at(self.db, a).endswith("+00:00"))
+
+    def test_a_case_variant_of_a_held_tag_is_not_added_and_names_stay_exact(self):
+        a = save(self.db, "Alpha", tags=["Portrait"])
+        b = save(self.db, "Beta")
+        result = self.run_bulk([a, b], ["portrait"])
+        # Alpha already holds it (casefold); Beta gets the name exactly as typed.
+        self.assertEqual((result["updated"], result["unchanged"]), (1, 1))
+        self.assertEqual(tags_of(self.db, a), ["Portrait"])
+        self.assertEqual(tags_of(self.db, b), ["portrait"])
+        self.assertEqual(tag_names(self.db), ["Portrait", "portrait"])
+        c = save(self.db, "Gamma", tags=["STRASSE"])
+        self.assertEqual(self.run_bulk([c], ["straße"])["unchanged"], 1)
+
+    def test_names_nothing_receives_are_never_inserted(self):
+        full = save(self.db, "Full", tags=[f"t{i}" for i in range(20)])
+        self.run_bulk([full, 99999], ["unused"])
+        self.assertNotIn("unused", tag_names(self.db))
+
+    def test_a_prompt_already_over_twenty_that_gains_nothing_is_unchanged(self):
+        over = save(self.db, "Over", tags=[f"t{i}" for i in range(22)])
+        result = self.run_bulk([over], ["t1"])
+        self.assertEqual((result["unchanged"], result["over_limit"]), (1, 0))
+
+    def test_image_and_pair_tags_are_untouched(self):
+        pair = self.db.save_prompt_pair("Cat", "Blur", tags=["pairtag"])
+        positive = self.db.get_prompt_pair(pair)["positive_prompt_id"]
+        with self.db.model.get_connection() as conn:
+            before = conn.execute("SELECT COUNT(*) FROM prompt_pair_tags").fetchone()[0]
+        self.run_bulk([positive], ["pairtag", "own"])
+        self.assertEqual(tags_of(self.db, positive), ["own", "pairtag"])
+        with self.db.model.get_connection() as conn:
+            after = conn.execute("SELECT COUNT(*) FROM prompt_pair_tags").fetchone()[0]
+        self.assertEqual(after, before)
+
+    def test_a_prompt_that_would_pass_twenty_tags_is_skipped_untouched(self):
+        full = save(self.db, "Full", tags=[f"t{i}" for i in range(19)])
+        other = save(self.db, "Other")
+        before = updated_at(self.db, full)
+        result = self.run_bulk([full, other], ["x", "y"])
+        self.assertEqual((result["updated"], result["over_limit"]), (1, 1))
+        self.assertEqual(len(tags_of(self.db, full)), 19)
+        self.assertEqual(updated_at(self.db, full), before)
+        self.assertEqual(tags_of(self.db, other), ["x", "y"])
+        # Exactly twenty is allowed.
+        self.assertEqual(self.run_bulk([full], ["x"])["updated"], 1)
+
+    def test_changed_prompts_become_registered_library_parents(self):
+        a = save(self.db, "Alpha")
+        self.run_bulk([a], ["new"])
+        with self.db.model.get_connection() as conn:
+            self.assertIsNotNone(
+                conn.execute(
+                    "SELECT 1 FROM prompt_library_state WHERE prompt_id=?", (a,)
+                ).fetchone()
+            )
+
+    def test_negative_only_prompts_are_not_found(self):
+        pair = self.db.save_prompt_pair("Cat", "Blur")
+        negative = self.db.get_prompt_pair(pair)["negative_prompt_id"]
+        result = self.run_bulk([negative], ["x"])
+        self.assertEqual((result["updated"], result["not_found"]), (0, 1))
+        self.assertEqual(tags_of(self.db, negative), [])
+
+    def test_restored_storage_writes_nothing_not_even_a_tag_name(self):
+        a = save(self.db, "Alpha")
+        rotate_history_epoch(self.db.model.get_connection())
+        with self.assertRaises(LibraryCommandError) as caught:
+            self.run_bulk([a], ["fresh"])
+        self.assertEqual(caught.exception.code, "storage-changed")
+        self.assertEqual(tags_of(self.db, a), [])
+        self.assertNotIn("fresh", tag_names(self.db))
+
+    def test_a_failure_mid_write_leaves_every_prompt_as_it_was(self):
+        ids = [save(self.db, f"P{i}") for i in range(3)]
+        with self.db.model.get_connection() as conn:
+            conn.execute(
+                "CREATE TRIGGER fail_third BEFORE INSERT ON prompt_tags "
+                f"WHEN NEW.prompt_id={ids[2]} BEGIN SELECT RAISE(ABORT, 'boom'); END"
+            )
+        with self.assertRaises(sqlite3.DatabaseError):
+            self.run_bulk(ids, ["x"])
+        self.assertEqual([tags_of(self.db, pid) for pid in ids], [[], [], []])
+        self.assertNotIn("x", tag_names(self.db))
+
+    def test_entries_are_counted_after_dedupe_and_the_limit_names_prompts(self):
+        a = save(self.db, "Alpha")
+        self.assertEqual(self.run_bulk([a], ["dup"] * 30 + ["Dup"])["updated"], 1)
+        with self.assertRaises(ValueError) as caught:
+            self.run_bulk([a], [f"t{i}" for i in range(101)])
+        self.assertEqual(str(caught.exception), "At most 20 tags can be added at once")
+
+    def test_many_ids_are_read_in_chunks(self):
+        ids = [save(self.db, f"Many {i}") for i in range(1000)]
+        self.assertEqual(self.run_bulk(ids, ["bulk"])["updated"], 1000)
+
+    def test_rejects_invalid_arguments(self):
+        a = save(self.db, "Alpha")
+        for ids, tags, epoch in (
+            ([a], [], self.epoch),
+            ([a], ["  "], self.epoch),
+            ([a], "red", self.epoch),
+            ([a], ["x" * 51], self.epoch),
+            ([a], ["bad\x01"], self.epoch),
+            ([a], ["a,b"], self.epoch),
+            ([a], [f"t{i}" for i in range(21)], self.epoch),
+            ([a], ["ok"], "not-a-uuid"),
+            ([], ["ok"], self.epoch),
+        ):
+            with self.subTest(tags=tags, epoch=epoch, ids=ids):
+                with self.assertRaises(ValueError):
+                    self.run_bulk(ids, tags, epoch)
+
+
+class TestBulkLibraryTagsRoute(RouteCoverageCase):
+    def epoch(self):
+        return self.api.db.get_library_options()["library_epoch"]
+
+    async def test_fenced_form_adds_tags_and_reports_counts(self):
+        ids = [self._save_prompt(f"route {i}") for i in range(2)]
+        status, body = await self._json(
+            "POST",
+            URL,
+            json={"prompt_ids": ids, "tags": ["Fresh"], "expected_epoch": self.epoch()},
+        )
+        self.assertEqual(status, 200)
+        self.assertEqual(
+            body,
+            {
+                "success": True,
+                "library_epoch": self.epoch(),
+                "tags": ["Fresh"],
+                "updated": 2,
+                "unchanged": 0,
+                "not_found": 0,
+                "over_limit": 0,
+            },
+        )
+
+    async def test_restored_storage_answers_409_storage_changed(self):
+        pid = self._save_prompt("route restore")
+        stale = self.epoch()
+        new_epoch = rotate_history_epoch(self.api.db.model.get_connection())
+        status, body = await self._json(
+            "POST",
+            URL,
+            json={"prompt_ids": [pid], "tags": ["x"], "expected_epoch": stale},
+        )
+        self.assertEqual((status, body["code"]), (409, "storage-changed"))
+        self.assertEqual(body["library_epoch"], new_epoch)
+
+    async def test_fenced_form_validates_its_body(self):
+        pid = self._save_prompt("route validate")
+        epoch = self.epoch()
+        for body in (
+            {"prompt_ids": [pid], "tags": [], "expected_epoch": epoch},
+            {"prompt_ids": [pid], "tags": "x", "expected_epoch": epoch},
+            {"prompt_ids": [pid], "tags": ["x" * 51], "expected_epoch": epoch},
+            {"prompt_ids": [pid], "tags": ["ok"], "expected_epoch": "nope"},
+            {"prompt_ids": [True], "tags": ["ok"], "expected_epoch": epoch},
+            {
+                "prompt_ids": list(range(1, 1002)),
+                "tags": ["ok"],
+                "expected_epoch": epoch,
+            },
+            {"prompt_ids": [pid], "expected_epoch": epoch},
+            {"prompt_ids": [pid], "tags": ["ok"], "expected_epoch": epoch, "x": 1},
+        ):
+            with self.subTest(body=body):
+                status, answer = await self._json("POST", URL, json=body)
+                self.assertEqual(status, 400)
+                self.assertFalse(answer["success"])
+        self.assertEqual(tags_of(self.api.db, pid), [])
+
+    async def test_fenced_form_reports_an_unexpected_failure(self):
+        pid = self._save_prompt("route failure")
+        self._break("bulk_add_library_tags")
+        status, body = await self._json(
+            "POST",
+            URL,
+            json={"prompt_ids": [pid], "tags": ["ok"], "expected_epoch": self.epoch()},
+        )
+        self.assertEqual(status, 500)
+        self.assertFalse(body["success"])
+
+    async def test_legacy_form_refuses_a_body_that_is_not_an_object(self):
+        for raw in ("[1, 2]", "not json"):
+            with self.subTest(raw=raw):
+                status, body = await self._json(
+                    "POST", URL, data=raw, headers={"Content-Type": "application/json"}
+                )
+                self.assertEqual(status, 400)
+                self.assertFalse(body["success"])
+
+    async def test_legacy_form_is_unchanged(self):
+        pid = self._save_prompt("route legacy")
+        status, body = await self._json(
+            "POST", URL, json={"prompt_ids": [pid], "tags": ["Legacy"]}
+        )
+        self.assertEqual((status, body["updated_count"]), (200, 1))
+        self.assertNotIn("library_epoch", body)
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_cached_runs.py

@@ -666,7 +666,7 @@ class TestOwnCounts(unittest.TestCase):
         return counts
 
 
-class TestReceipts(unittest.TestCase):
+class ReceiptCase(unittest.TestCase):
     def setUp(self):
         self.context = ("j1", None)
         self.runs = CachedRuns(
@@ -686,6 +686,8 @@ class TestReceipts(unittest.TestCase):
     def keys(self):
         return list(self.runs._receipts)
 
+
+class TestReceipts(ReceiptCase):
     def test_least_recently_used_run_keys_are_dropped(self):
         self.begin(1)
         for name in ("a", "b", "c"):
@@ -726,6 +728,196 @@ class TestReceipts(unittest.TestCase):
         self.assertEqual(list(self.runs._attempt.nodes), ["b"])
 
 
+class TestRecordedFields(ReceiptCase):
+    """The texts a receipt keeps for the history capture of a cached job (#403)."""
+
+    F = (("raw a", " pre ", ""),)
+
+    def record(self, item_id, name="a", output="x", fields=F, facts=None):
+        self.begin(item_id)
+        self.runs.record_item(name, (output,), facts or Facts("h"), fields)
+        self.runs.end_attempt(item_id)
+
+    def lookup(self, name="a", values=(("x",),), prompt_id="j9", item_id=9):
+        self.begin(item_id, prompt_id)
+        return self.runs.recorded_fields(prompt_id, name, values)
+
+    def test_a_cached_job_reads_the_fields_of_the_execution_it_reuses(self):
+        self.record(1)
+        self.assertEqual(self.lookup(), [self.F])
+
+    def test_list_items_return_their_own_fields_in_order(self):
+        self.begin(1)
+        for index, text in enumerate(("x", "y")):
+            self.context = ("j1", index)
+            self.runs.record_item("a", (text,), Facts("h"), ((text, "", ""),))
+        self.runs.end_attempt(1)
+        self.assertEqual(
+            self.lookup(values=(("x", "y"),)), [(("x", "", ""),), (("y", "", ""),)]
+        )
+
+    def test_fails_closed(self):
+        self.record(1)
+        for case, kwargs in (
+            ("other output", dict(values=(("changed",),))),
+            ("other node", dict(name="b")),
+            ("unknown node", dict(name="zzz")),
+            ("no values", dict(values=())),
+            ("malformed values", dict(values="x")),
+        ):
+            with self.subTest(case=case):
+                self.assertIsNone(self.lookup(**kwargs))
+        with self.subTest(case="another job"):
+            self.begin(9, "j9")
+            self.assertIsNone(self.runs.recorded_fields("j8", "a", (("x",),)))
+        with self.subTest(case="no job"):
+            self.runs.end_attempt(9)
+            self.assertIsNone(self.runs.recorded_fields("j9", "a", (("x",),)))
+
+    def test_an_execution_without_fields_gives_none(self):
+        for fields in (None, (("a", "b"),), ((1, "", ""),), (), "abc"):
+            with self.subTest(fields=fields):
+                self.runs = CachedRuns(context=lambda: self.context)
+                self.record(1, fields=fields)
+                self.assertIsNone(self.lookup())
+
+    def test_a_lookup_never_counts_or_marks_the_node_replayed(self):
+        self.record(1)
+        self.lookup()
+        self.assertFalse(self.runs.replayed("a"))
+
+    def test_equal_outputs_with_other_raw_text_conflict_and_still_count(self):
+        key = prompt_node_key(self.g, "a")
+        self.record(1, fields=(("x ", "", ""),))
+        self.record(2, fields=(("x", "", ""),))
+        # Counting dedupes by outputs as before; the fields cannot name either
+        self.assertEqual([r.item_id for r in self.runs._receipts[key]], [2])
+        self.assertIsNone(self.lookup())
+        self.assertIsNotNone(self.runs._match(key, (("x",),)))
+        # The conflict stays: a later execution cannot tell which text it was
+        self.record(3, fields=(("x", "", ""),))
+        self.assertIsNone(self.lookup())
+
+    def test_equal_outputs_with_other_facts_still_count_the_newest(self):
+        key = prompt_node_key(self.g, "a")
+        self.record(1, facts=Facts("old"), fields=(("a x", "", ""),))
+        self.record(2, facts=Facts("new"), fields=(("x", "a ", ""),))
+        items = self.runs._match(key, (("x",),))
+        self.assertEqual([item.facts for item in items], [Facts("new")])
+
+    def test_equal_outputs_and_fields_still_replace(self):
+        key = prompt_node_key(self.g, "a")
+        self.record(1)
+        self.record(2)
+        self.assertEqual([r.item_id for r in self.runs._receipts[key]], [2])
+        self.assertEqual(self.lookup(), [self.F])
+
+    def test_fields_dropped_by_the_budget_do_not_conflict(self):
+        self.runs = CachedRuns(context=lambda: self.context, max_field_bytes=0)
+        self.record(1)
+        self.runs._max_field_bytes = 100
+        self.record(2)
+        self.assertEqual(self.lookup(), [self.F])
+
+    def test_oversized_or_invalid_fields_are_not_kept(self):
+        for raw in ("x" * (module.MAX_FIELD_BYTES + 1), "\ud800"):
+            with self.subTest(size=len(raw)):
+                self.runs = CachedRuns(context=lambda: self.context)
+                self.record(1, fields=((raw, "", ""),))
+                self.assertIsNone(self.lookup())
+                self.assertEqual(self.runs._field_bytes, 0)
+
+    def test_old_receipts_give_up_their_fields_to_new_ones(self):
+        self.runs = CachedRuns(
+            context=lambda: self.context, max_keys=3, max_field_bytes=10
+        )
+        self.record(1, "a", fields=(("123456", "", ""),))
+        self.record(2, "b", fields=(("123456", "", ""),))
+        self.assertEqual(self.runs._field_bytes, 6)
+        self.assertIsNone(self.lookup("a"))
+        self.assertEqual(self.lookup("b"), [(("123456", "", ""),)])
+        # "a" still counts
+        self.assertIsNotNone(self.runs._match(prompt_node_key(self.g, "a"), (("x",),)))
+        # Larger than the whole budget: not kept, nothing else given up
+        self.record(3, "c", fields=(("12345678901", "", ""),))
+        self.assertEqual(self.runs._field_bytes, 6)
+        self.assertEqual(self.lookup("b"), [(("123456", "", ""),)])
+
+    def test_one_receipt_holds_at_most_one_runs_history(self):
+        with mock.patch.object(module, "MAX_RECEIPT_FIELD_BYTES", 8):
+            self.begin(1)
+            for index, raw in enumerate(("12345", "12345")):
+                self.context = ("j1", index)
+                self.runs.record_item("a", (raw,), Facts("h"), ((raw, "", ""),))
+            self.runs.end_attempt(1)
+        self.assertEqual(self.runs._field_bytes, 5)
+        self.assertIsNone(self.lookup(values=(("12345", "12345"),)))
+
+    def test_the_byte_count_follows_replacement_and_eviction(self):
+        self.begin(1)
+        self.runs.record_item("a", ("x",), Facts("h"), (("12345", "", ""),))
+        self.runs.record_item("a", ("x",), Facts("h"), (("123", "", ""),))
+        self.assertEqual(self.runs._field_bytes, 3)
+        self.runs.end_attempt(1)
+        for item_id, raw in ((2, "1"), (3, "22"), (4, "333")):
+            self.record(item_id, output=raw, fields=((raw, "", ""),))
+        # max_receipts=2: only the receipts of jobs 4 and 3 remain
+        self.assertEqual(self.runs._field_bytes, 5)
+        for item_id, name in ((5, "b"), (6, "c")):  # max_keys=2: "a" is dropped
+            self.record(item_id, name, fields=(("1", "", ""),))
+        self.assertEqual(self.runs._field_bytes, 2)
+        self.assertEqual(
+            self.runs._field_bytes,
+            sum(module._bytes(r) for r in self.runs._receipts.values()),
+        )
+
+
+class TestCachedJobsRecordFields(JobCase):
+    """The real nodes keep the texts they gave the history capture (#403)."""
+
+    def lookup(self, g, node_id="pm", prompt_id="job-x"):
+        self.item_id += 1
+        self.runs.begin_attempt(
+            ((self.item_id, prompt_id, g, {}, ["save"], {}), self.item_id)
+        )
+        values = cached_string_outputs(self.executor, node_id, g[node_id]["class_type"])
+        found = self.runs.recorded_fields(prompt_id, node_id, values)
+        self.runs.end_attempt(self.item_id)
+        return found
+
+    def test_each_node_kind_keeps_its_runtime_fields(self):
+        for kind in (
+            "PromptManager",
+            "PromptManagerFocused",
+            "PromptManagerSplit",
+            "PromptManagerText",
+        ):
+            with self.subTest(kind=kind):
+                g = graph(kind)
+                self.first(g, text=f" {LLM} ")
+                self.assertEqual(
+                    self.lookup(graph(kind, seed=2)), [((f" {LLM} ", "", ""),)]
+                )
+
+    def test_pair_keeps_positive_and_negative_fields(self):
+        g = graph("PromptManagerPair", negative_text="typed negative")
+        self.first(g, negative="typed negative")
+        self.assertEqual(
+            self.lookup(
+                graph("PromptManagerPair", negative_text="typed negative", seed=2)
+            ),
+            [((LLM, "", ""), ("typed negative", "", ""))],
+        )
+
+    def test_a_disabled_node_keeps_no_fields(self):
+        g = graph()
+        g["pm"]["inputs"]["save_to_library"] = False
+        self.first(g)
+        g2 = graph(seed=2)
+        g2["pm"]["inputs"]["save_to_library"] = False
+        self.assertIsNone(self.lookup(g2))
+
+
 class TestCacheReader(unittest.TestCase):
     def setUp(self):
         patcher = mock.patch.dict(sys.modules, host_modules())

tests/test_execution_context_host.py

@@ -100,14 +100,19 @@ class HostTests(unittest.TestCase):
         self.executor.caches = types.SimpleNamespace(outputs=HierarchicalCache())
 
     def begin(
-        self, kind="PromptManager", raw=" raw ", prepend=" pre ", append=" post "
+        self,
+        kind="PromptManager",
+        raw=" raw ",
+        prepend=" pre ",
+        append=" post ",
+        negative="negative",
     ):
         g = {
             "p": {
                 "class_type": kind,
                 "inputs": {
                     "text": raw,
-                    "negative_text": "negative",
+                    "negative_text": negative,
                     "prepend_text": prepend,
                     "append_text": append,
                 },
@@ -125,7 +130,7 @@ class HostTests(unittest.TestCase):
                 },
             },
         }
-        if type(raw) is list or type(prepend) is list:
+        if any(type(value) is list for value in (raw, prepend, negative)):
             g["dynamic"] = {"class_type": "RuntimeGenerator", "inputs": {}}
             g["stateful"] = {
                 "class_type": "PromptManagerText",
@@ -445,3 +450,138 @@ class HostTests(unittest.TestCase):
         ref = install_and_remove()
         gc.collect()
         self.assertIsNone(ref())
+
+    # -- #403: a cached runtime text takes its fields from the reused execution
+
+    class Receipts:
+        """A fake ``CachedRuns.recorded_fields``: fixed fields, calls recorded."""
+
+        def __init__(self, fields):
+            self.fields, self.calls = fields, []
+
+        def recorded_fields(self, prompt_id, node_id, values):
+            self.calls.append((prompt_id, node_id, values))
+            return self.fields
+
+    def observe_with(self, receipts, outputs, **begin):
+        self.begin(**begin)
+        self.executor.caches.outputs.entries["p"] = Entry(None, outputs)
+        self.observe("execution_start")
+        host.observe_cached_context(
+            self.executor,
+            "execution_cached",
+            {"prompt_id": JOB, "nodes": ["p"]},
+            receipts=receipts,
+        )
+        return self.seal()
+
+    def test_cached_runtime_raw_comes_from_the_matching_receipt(self):
+        receipts = self.Receipts([(("LLM raw", " pre ", " post "),)])
+        result = self.observe_with(
+            receipts, [object(), ["cached"]], raw=["dynamic", 0]
+        ).positive
+        self.assertEqual(receipts.calls, [(JOB, "p", (("cached",),))])
+        self.assertEqual(result.raw.value, "LLM raw")
+        self.assertEqual(
+            (result.raw.provenance, result.raw.source), ("inferred", "graph_semantics")
+        )
+        self.assertEqual(result.effective.value, "cached")
+        self.assertEqual(result.effective.source, "host_cache_output")
+        # Fields the graph already knows are never replaced
+        self.assertEqual(result.prepend.value, " pre ")
+        self.assertEqual(result.prepend.source, "graph_literal")
+        self.assertEqual(result.append.value, " post ")
+
+    def test_cached_runtime_prepend_comes_from_the_matching_receipt(self):
+        receipts = self.Receipts([((" raw ", "LLM pre", " post "),)])
+        result = self.observe_with(
+            receipts, [object(), ["cached"]], prepend=["dynamic", 0]
+        ).positive
+        self.assertEqual(result.prepend.value, "LLM pre")
+        self.assertEqual(result.prepend.source, "graph_semantics")
+        self.assertEqual(result.raw.value, " raw ")
+        self.assertEqual(result.raw.source, "graph_literal")
+
+    def test_a_receipt_disagreeing_with_a_known_field_fills_nothing(self):
+        receipts = self.Receipts([(("LLM raw", "other pre", " post "),)])
+        result = self.observe_with(
+            receipts, [object(), ["cached"]], raw=["dynamic", 0]
+        ).positive
+        self.assertIsNone(result.raw.value)
+        self.assertEqual(result.prepend.value, " pre ")
+
+    def test_pair_runtime_negative_takes_the_negative_fields(self):
+        receipts = self.Receipts(
+            [((" raw ", " pre ", " post "), ("LLM negative", "", ""))]
+        )
+        result = self.observe_with(
+            receipts,
+            [object(), object(), ["cached positive"], ["LLM negative"]],
+            kind="PromptManagerPair",
+            negative=["dynamic", 0],
+        )
+        self.assertEqual(result.negative.raw.value, "LLM negative")
+        self.assertEqual(result.negative.raw.source, "graph_semantics")
+        self.assertEqual(result.positive.raw.value, " raw ")
+        self.assertEqual(result.positive.raw.source, "graph_literal")
+
+    def test_known_typed_fields_are_never_replaced_by_a_receipt(self):
+        receipts = self.Receipts([(("other", "other", "other"),)])
+        result = self.observe_with(receipts, [object(), ["cached"]]).positive
+        self.assertEqual(
+            (result.raw.value, result.prepend.value, result.append.value),
+            (" raw ", " pre ", " post "),
+        )
+        self.assertEqual(result.raw.source, "graph_literal")
+
+    def test_no_matching_receipt_keeps_the_cached_raw_unknown(self):
+        for fields in (None, [None], [], "bad", [(("a", "b"),)], [((1, "", ""),)]):
+            with self.subTest(fields=fields):
+                result = self.observe_with(
+                    self.Receipts(fields), [object(), ["cached"]], raw=["dynamic", 0]
+                ).positive
+                self.assertIsNone(result.raw.value)
+                self.assertEqual(result.effective.value, "cached")
+
+    def test_a_failing_receipt_lookup_keeps_the_capture(self):
+        class Broken:
+            def recorded_fields(self, *args):
+                raise RuntimeError("boom")
+
+        result = self.observe_with(
+            Broken(), [object(), ["cached"]], raw=["dynamic", 0]
+        ).positive
+        self.assertIsNone(result.raw.value)
+        self.assertEqual(result.effective.value, "cached")
+
+    def test_pair_slots_take_their_own_receipt_fields(self):
+        receipts = self.Receipts(
+            [(("LLM positive", " pre ", " post "), ("LLM negative", "", ""))]
+        )
+        result = self.observe_with(
+            receipts,
+            [object(), object(), ["cached positive"], ["cached negative"]],
+            kind="PromptManagerPair",
+            raw=["dynamic", 0],
+        )
+        self.assertEqual(
+            receipts.calls[0], (JOB, "p", (("cached positive",), ("cached negative",)))
+        )
+        self.assertEqual(result.positive.raw.value, "LLM positive")
+        self.assertEqual(result.positive.effective.value, "cached positive")
+        # The negative editor is typed: the graph's value stays
+        self.assertEqual(result.negative.raw.value, "negative")
+        self.assertEqual(result.negative.raw.source, "graph_literal")
+
+    def test_default_receipts_are_the_process_cached_runs(self):
+        from utils import cached_runs
+
+        receipts = self.Receipts([(("LLM raw", " pre ", " post "),)])
+        with patch.object(cached_runs, "CACHED_RUNS", receipts):
+            self.begin(raw=["dynamic", 0])
+            self.executor.caches.outputs.entries["p"] = Entry(
+                None, [object(), ["cached"]]
+            )
+            self.observe("execution_start")
+            self.observe("execution_cached", nodes=["p"])
+        self.assertEqual(self.seal().positive.raw.value, "LLM raw")

tests/test_execution_history_hooks.py

@@ -54,7 +54,7 @@ def status(kind="execution_success", job=JOB):
     return Status("success", True, [(kind, {"prompt_id": job, "timestamp": 20})])
 
 
-class HistoryHooksTests(unittest.TestCase):
+class HooksCase(unittest.TestCase):
     def setUp(self):
         self.directory = tempfile.TemporaryDirectory()
         self.addCleanup(self.directory.cleanup)
@@ -122,6 +122,8 @@ class HistoryHooksTests(unittest.TestCase):
     def drain(self, registration):
         drain_writer(registration.writer, clock=time.monotonic)
 
+
+class HistoryHooksTests(HooksCase):
     def test_package_history_initializes_when_monitor_and_usage_setup_fail(self):
         try:
             from tests.test_init_package import (
@@ -767,3 +769,150 @@ class HistoryHooksTests(unittest.TestCase):
         self.assertEqual(
             calls, ["execution_start", "execution_cached", "execution_cached"]
         )
+
+
+class CachedRuntimeHistoryTests(HooksCase):
+    """A cached runtime-linked prompt is a tracked run, like an executed one (#403).
+
+    The real history hook and writer, a real ``CachedRuns`` and the real host
+    adapter over fakes of ComfyUI's executor, cache and executing context.
+    """
+
+    TEXT = "a lighthouse the LLM wrote"
+    GRAPH = {
+        "str": {"class_type": "PrimitiveString", "inputs": {"value": "seed"}},
+        "llm": {"class_type": "PreviewAny", "inputs": {"source": ["str", 0]}},
+        "p": {"class_type": "PromptManager", "inputs": {"text": ["llm", 0]}},
+        "s": {"class_type": "KSampler", "inputs": {"positive": ["p", 0]}},
+    }
+
+    def setUp(self):
+        super().setUp()
+        from utils import cached_runs
+        from utils.cached_runs import CachedRuns
+
+        self.context = None
+        self.runs = CachedRuns(context=lambda: self.context or (None, None))
+        self.cache_entries = {}
+        entries = self.cache_entries
+        execution = types.ModuleType("execution")
+        caching = types.ModuleType("comfy_execution.caching")
+        package = types.ModuleType("comfy_execution")
+        utils_module = types.ModuleType("comfy_execution.utils")
+        entry_type = namedtuple("CacheEntry", "ui outputs")
+
+        class PromptExecutor:
+            def add_message(self, event, data, broadcast):
+                return None
+
+        class HierarchicalCache:
+            def get_local(self, node_id):
+                return entries.get(node_id)
+
+        execution.PromptExecutor, execution.CacheEntry = PromptExecutor, entry_type
+        caching.HierarchicalCache = HierarchicalCache
+        package.caching, package.utils = caching, utils_module
+        utils_module.get_executing_context = lambda: (
+            SimpleNamespace(prompt_id=self.context[0], node_id="p", list_index=0)
+            if self.context
+            else None
+        )
+        self.entry_type = entry_type
+        modules = patch.dict(
+            sys.modules,
+            {
+                "execution": execution,
+                "comfy_execution": package,
+                "comfy_execution.caching": caching,
+                "comfy_execution.utils": utils_module,
+            },
+        )
+        modules.start()
+        self.addCleanup(modules.stop)
+        swap = patch.object(cached_runs, "CACHED_RUNS", self.runs)
+        swap.start()
+        self.addCleanup(swap.stop)
+        self.executor = PromptExecutor()
+        self.executor.server = self.server
+        self.executor.caches = SimpleNamespace(outputs=HierarchicalCache())
+        self.prompt_id = self.db.save_prompt(text=self.TEXT)
+        self.item_id = 0
+
+    def job(self, cached):
+        """One job; the node executes unless ``cached`` (ComfyUI's cache serves it)."""
+        from uuid import uuid4
+
+        from utils.cached_runs import stored_facts
+        from utils.execution_context_capture import record_runtime_context
+
+        self.item_id += 1
+        job = str(uuid4())
+        result = ((0, job, self.GRAPH, {}, ["s"]), self.item_id)
+        self.queue.results.append(result)
+        self.queue.get()
+        self.runs.begin_attempt(result)
+        self.context = (job, 0)
+        self.executor.add_message("execution_start", {"prompt_id": job}, False)
+        if cached:
+            self.executor.add_message(
+                "execution_cached", {"prompt_id": job, "nodes": ["p"]}, False
+            )
+        else:
+            # What PromptManager.encode_prompt records for this text
+            record_runtime_context("PromptManager", self.TEXT, "", "", self.TEXT)
+            self.runs.record_item(
+                "p", (self.TEXT,), stored_facts(self.TEXT), ((self.TEXT, "", ""),)
+            )
+            self.cache_entries["p"] = self.entry_type(None, [object(), [self.TEXT]])
+        self.context = None
+        self.queue.task_done(self.item_id, {}, status(job=job))
+        self.runs.end_attempt(self.item_id)
+
+    def tracked_runs(self):
+        with self.db.recorded_history_snapshot():
+            return self.db.list_history_window(self.prompt_id)["recorded_run_count"]
+
+    def test_three_jobs_of_runtime_text_are_three_tracked_runs(self):
+        registration = self.register()
+        for cached in (False, True, True):
+            self.job(cached)
+        self.drain(registration)
+        self.assertEqual(self.tracked_runs(), 3)
+        rows = self.rows()
+        for row in rows:
+            contexts = self.db.get_execution_contexts(row["run_id"])["contexts"]
+            self.assertEqual(contexts[0]["positive"]["raw"], self.TEXT)
+            self.assertEqual(contexts[0]["positive_prompt_id"], self.prompt_id)
+        cached = self.db.get_execution_contexts(rows[1]["run_id"])["contexts"][0]
+        self.assertEqual(
+            cached["evidence"]["positive"]["raw"],
+            {"provenance": "inferred", "source": "graph_semantics", "reason": None},
+        )
+        self.assertEqual(
+            cached["evidence"]["positive"]["effective"]["source"], "host_cache_output"
+        )
+        # The stored history passes the whole-database check older builds run too
+        from database.execution_history_schema import history_schema_error
+
+        self.assertIsNone(history_schema_error(self.db.model.get_connection()))
+        # An image of a cached job can now be attributed through its run (#342)
+        from database.run_attribution import named_prompt
+
+        self.assertEqual(
+            [named_prompt(self.db, row["run_id"]) for row in rows],
+            [self.prompt_id] * 3,
+        )
+
+    def test_without_a_receipt_a_cached_job_stays_untracked(self):
+        registration = self.register()
+        self.job(False)
+        self.runs._receipts.clear()  # as after a restart with a provider entry
+        self.job(True)
+        self.drain(registration)
+        self.assertEqual(self.tracked_runs(), 1)
+        cached = self.db.get_execution_contexts(self.rows()[1]["run_id"])["contexts"]
+        self.assertIsNone(cached[0]["positive"]["raw"])
+        self.assertIsNone(cached[0]["positive_prompt_id"])
+        from database.run_attribution import named_prompt
+
+        self.assertIsNone(named_prompt(self.db, self.rows()[1]["run_id"]))

tests/test_pair_usage_tracking.py

@@ -99,7 +99,10 @@ class TestPairUsageTracking(unittest.TestCase):
             unique_id="1",
         )
         node.cached_runs.record_item.assert_called_once_with(
-            "1", ("cat", ""), Facts(usage_tracking.generate_prompt_hash("cat"), None)
+            "1",
+            ("cat", ""),
+            Facts(usage_tracking.generate_prompt_hash("cat"), None),
+            (("cat", "", ""), ("", "", "")),
         )
 
     def test_validation_rejection_never_counts_or_saves(self):

tests/test_prompt_usage.py

@@ -308,7 +308,11 @@ class TestNodeRecordsCachedRunReceipt(UsageTestCase):
             "typed text", prepend_text="pre", prompt=self.GRAPH, unique_id="1"
         )
         node.cached_runs.record_item.assert_called_once_with(
-            "1", ("pre typed text",), Facts(generate_prompt_hash("typed text"))
+            "1",
+            ("pre typed text",),
+            Facts(generate_prompt_hash("typed text")),
+            # #403: the texts the history capture got, for a cached job's history
+            (("typed text", "pre", ""),),
         )
         # Hook-visible text: the node counts nothing itself
         row = self.db.get_prompt_by_hash(generate_prompt_hash("typed text"))
@@ -319,7 +323,9 @@ class TestNodeRecordsCachedRunReceipt(UsageTestCase):
 
         node = self._node()
         node.process_text("  ", prompt=self.GRAPH, unique_id="1")
-        node.cached_runs.record_item.assert_called_once_with("1", ("  ",), Facts(None))
+        node.cached_runs.record_item.assert_called_once_with(
+            "1", ("  ",), Facts(None), (("  ", "", ""),)
+        )
 
 
 if __name__ == "__main__":

tools/vendor/frontend-coverage-baseline.json

@@ -275,6 +275,12 @@
     "functions": 98,
     "statements": 98
   },
+  "workspace/images-collections.mjs": {
+    "lines": 98,
+    "branches": 89,
+    "functions": 98,
+    "statements": 98
+  },
   "workspace/images-export.mjs": {
     "lines": 98,
     "branches": 98,
@@ -353,6 +359,12 @@
     "functions": 98,
     "statements": 98
   },
+  "workspace/inspector-collections.mjs": {
+    "lines": 98,
+    "branches": 92,
+    "functions": 98,
+    "statements": 98
+  },
   "workspace/library-editor-state.mjs": {
     "lines": 96,
     "branches": 92,
@@ -377,6 +389,12 @@
     "functions": 93,
     "statements": 94
   },
+  "workspace/library-selection.mjs": {
+    "lines": 98,
+    "branches": 86,
+    "functions": 84,
+    "statements": 98
+  },
   "workspace/library-state.mjs": {
     "lines": 98,
     "branches": 95,

utils/cached_runs.py

@@ -15,9 +15,14 @@ role (positive, negative, which Pair outputs are used) and what the hook
 already saw are recomputed from that job's own graph, and both paths count
 through the same ``own_counts`` so they cannot drift.
 
+The receipt also keeps the raw, prepend and append text the node gave the
+execution history (#403), so the history capture of a cached job can record
+the raw text the graph cannot show (``recorded_fields``) and the job becomes a
+tracked run of its prompt, as it is a counted one.
+
 Fail closed: when the cached text cannot be matched to a receipt recorded in
-this process, nothing is counted (the behavior before #347). In memory only and
-bounded; starts no threads.
+this process, nothing is counted (the behavior before #347) and the history
+fields stay unknown. In memory only and bounded; starts no threads.
 """
 
 import hashlib
@@ -39,6 +44,7 @@ try:
     from .queue_observers import subscribe_dequeue, subscribe_task_done
     from .run_keys import prompt_node_key, upstream_nodes
     from .usage_tracking import find_pair_id_by_hashes, survive_switch
+    from .execution_contexts import MAX_ATTEMPT_BYTES, MAX_FIELD_BYTES, text_size
 except ImportError:
     from utils.hashing import generate_prompt_hash
     from utils.logging_config import get_logger
@@ -53,6 +59,7 @@ except ImportError:
     from utils.queue_observers import subscribe_dequeue, subscribe_task_done
     from utils.run_keys import prompt_node_key, upstream_nodes
     from utils.usage_tracking import find_pair_id_by_hashes, survive_switch
+    from utils.execution_contexts import MAX_ATTEMPT_BYTES, MAX_FIELD_BYTES, text_size
 
 logger = get_logger("prompt_manager.cached_runs")
 
@@ -61,9 +68,15 @@ MAX_KEYS = 256  # run keys remembered (the same bound as MAX_KEYED_RUNS)
 MAX_RECEIPTS = 4  # receipts per run key: a recurring upstream value (A, B, A)
 MAX_ITEMS = 64  # list items per receipt (the cache reader's bound)
 MAX_CACHED_NODES = 500  # the execution_cached guard the history capture uses
+MAX_FIELD_TEXT_BYTES = 8 * 1024 * 1024  # history fields all receipts keep
+MAX_RECEIPT_FIELD_BYTES = MAX_ATTEMPT_BYTES  # one receipt: what one run's history holds
 _EMPTY_HASH = generate_prompt_hash("")
 
 Count = Tuple[Any, ...]  # ("prompt", hash) or ("pair", positive, negative or None)
+# (raw, prepend, append) per STRING slot, as given to record_runtime_context
+Fields = Tuple[Tuple[str, str, str], ...]
+# Executions with equal outputs recorded different fields: none can be named
+_CONFLICT: Fields = ()
 
 
 class Facts(NamedTuple):
@@ -96,6 +109,8 @@ class Seen(NamedTuple):
 class _Item(NamedTuple):
     outputs: Tuple[str, ...]
     facts: Facts
+    fields: Optional[Fields] = None
+    field_bytes: int = 0
 
 
 class _Receipt:
@@ -108,6 +123,18 @@ class _Receipt:
     def outputs(self) -> Dict[int, Tuple[str, ...]]:
         return {index: item.outputs for index, item in self.items.items()}
 
+    def field_bytes(self) -> int:
+        return sum(item.field_bytes for item in self.items.values())
+
+    def strip_fields(self) -> int:
+        """Drop the history fields (counting is unchanged); the bytes freed."""
+        freed = 0
+        for index, item in list(self.items.items()):
+            if item.field_bytes:
+                freed += item.field_bytes
+                self.items[index] = item._replace(fields=None, field_bytes=0)
+        return freed
+
 
 class _Attempt(NamedTuple):
     item_id: int
@@ -117,6 +144,22 @@ class _Attempt(NamedTuple):
     replayed: set
 
 
+def _field_size(fields: Any, slots: int) -> Optional[int]:
+    """UTF-8 size of well-formed fields for ``slots`` outputs, else None."""
+    if type(fields) is not tuple or len(fields) != slots:
+        return None
+    size = 0
+    for slot in fields:
+        if type(slot) is not tuple or len(slot) != 3:
+            return None
+        for value in slot:
+            try:
+                size += text_size(value, MAX_FIELD_BYTES)  # the capture's own check
+            except ValueError:
+                return None
+    return size
+
+
 def text_digest(value: str) -> str:
     """Digest of one returned STRING value, exactly as returned."""
     return hashlib.sha256(value.encode("utf-8", "surrogatepass")).hexdigest()
@@ -245,11 +288,14 @@ class CachedRuns:
         max_keys: int = MAX_KEYS,
         max_receipts: int = MAX_RECEIPTS,
         max_items: int = MAX_ITEMS,
+        max_field_bytes: int = MAX_FIELD_TEXT_BYTES,
         context: Callable[[], Tuple[Optional[str], Optional[int]]] = _host_context,
     ) -> None:
         self._max_keys = max_keys
         self._max_receipts = max_receipts
         self._max_items = max_items
+        self._max_field_bytes = max_field_bytes
+        self._field_bytes = 0  # UTF-8 bytes of the fields all receipts keep
         self._context = context
         self._receipts: "OrderedDict[str, List[_Receipt]]" = OrderedDict()
         self._attempt: Optional[_Attempt] = None
@@ -284,7 +330,10 @@ class CachedRuns:
                 self._attempt = None
             for run_key, receipts in self._receipts.items():
                 if receipts and receipts[0].item_id == item_id:
-                    self._receipts[run_key] = _deduplicated(receipts)
+                    before = _bytes(receipts)
+                    deduplicated = _deduplicated(receipts)
+                    self._field_bytes += _bytes(deduplicated) - before
+                    self._receipts[run_key] = deduplicated
 
     def _current(self) -> Tuple[Optional[_Attempt], Optional[int]]:
         """The attempt the executing node belongs to, and its list index."""
@@ -305,8 +354,19 @@ class CachedRuns:
         with self._lock:
             return str(node_id) in attempt.replayed
 
-    def record_item(self, node_id: Any, outputs: Tuple[Any, ...], facts: Facts) -> None:
-        """Remember what this execution returned and was entitled to count."""
+    def record_item(
+        self,
+        node_id: Any,
+        outputs: Tuple[Any, ...],
+        facts: Facts,
+        fields: Optional[Fields] = None,
+    ) -> None:
+        """Remember what this execution returned and was entitled to count.
+
+        ``fields`` are the texts the node gave the execution history, one
+        ``(raw, prepend, append)`` per output; kept only when well formed,
+        each within the history's field limit and the store within its budget.
+        """
         try:
             attempt, index = self._current()
             node = attempt.nodes.get(str(node_id)) if attempt else None
@@ -317,14 +377,20 @@ class CachedRuns:
                 return
             if any(not isinstance(value, str) for value in outputs):
                 return
-            item = _Item(tuple(text_digest(value) for value in outputs), facts)
+            size = _field_size(fields, len(outputs)) if fields is not None else None
+            item = _Item(
+                tuple(text_digest(value) for value in outputs),
+                facts,
+                fields if size is not None else None,
+                size or 0,
+            )
             self._store(node[1], attempt.item_id, index, item)
         except Exception as error:
             logger.debug(f"Could not record a cached-run receipt: {error}")
 
     def _store(self, run_key: str, item_id: int, index: int, item: _Item) -> None:
         with self._lock:
-            receipts = self._receipts.pop(run_key, [])
+            receipts = self._take(run_key)
             if receipts and receipts[0].item_id == item_id:
                 receipt = receipts[0]
             else:
@@ -334,9 +400,44 @@ class CachedRuns:
                 receipt = _Receipt(item_id)
                 receipts.insert(0, receipt)
             receipt.items[index] = item
-            self._receipts[run_key] = receipts[: self._max_receipts]
+            receipts = receipts[: self._max_receipts]
+            if item.field_bytes and (
+                receipt.field_bytes() > MAX_RECEIPT_FIELD_BYTES
+                or not self._make_room(_bytes(receipts))
+            ):
+                # Counting is as before; this job's history fields stay unknown
+                receipt.items[index] = item._replace(fields=None, field_bytes=0)
+            self._put(run_key, receipts)
             while len(self._receipts) > self._max_keys:
-                self._receipts.popitem(last=False)
+                _, dropped = self._receipts.popitem(last=False)
+                self._field_bytes -= _bytes(dropped)
+
+    def _make_room(self, needed: int) -> bool:
+        """Free the fields of the least recently used receipts (under the lock).
+
+        Old receipts of workflows never run again must not hold the budget:
+        their counting stays, only their history fields go. False when
+        ``needed`` does not fit even then.
+        """
+        if needed > self._max_field_bytes:
+            return False
+        for receipts in self._receipts.values():  # least recently used first
+            if self._field_bytes + needed <= self._max_field_bytes:
+                break
+            for receipt in receipts:
+                self._field_bytes -= receipt.strip_fields()
+        return self._field_bytes + needed <= self._max_field_bytes
+
+    def _take(self, run_key: str) -> List[_Receipt]:
+        """Remove a key's receipts and their bytes from the store (under the lock)."""
+        receipts = self._receipts.pop(run_key, [])
+        self._field_bytes -= _bytes(receipts)
+        return list(receipts)
+
+    def _put(self, run_key: str, receipts: List[_Receipt]) -> None:
+        """Store a key's receipts as the most recently used (under the lock)."""
+        self._receipts[run_key] = receipts
+        self._field_bytes += _bytes(receipts)
 
     # -- cached nodes -------------------------------------------------------
 
@@ -393,25 +494,11 @@ class CachedRuns:
 
     def _match(self, run_key: str, values: Any) -> Optional[List[_Item]]:
         """The items of the one receipt whose outputs equal ``values``, in order."""
-        if not values:
-            return None
-        length = len(values[0])
-        if length == 0 or length > self._max_items:
-            return None
-        if any(len(slot) != length for slot in values):
-            return None
-        wanted = {
-            index: tuple(text_digest(slot[index]) for slot in values)
-            for index in range(length)
-        }
         with self._lock:
-            receipts = self._receipts.get(run_key)
-            if not receipts:
-                return None
-            matches = [receipt for receipt in receipts if receipt.outputs() == wanted]
+            matches = self._matching(run_key, values)
             if not matches:
                 return None
-            self._receipts.move_to_end(run_key)
+            length = len(values[0])
             facts = {
                 tuple(receipt.items[index].facts for index in range(length))
                 for receipt in matches
@@ -420,13 +507,99 @@ class CachedRuns:
                 return None
             return [matches[0].items[index] for index in range(length)]
 
+    def _matching(self, run_key: str, values: Any) -> List[_Receipt]:
+        """Receipts of ``run_key`` whose outputs equal ``values`` (under the lock)."""
+        if type(values) not in (list, tuple) or not values:
+            return []
+        if any(type(slot) not in (list, tuple) for slot in values):
+            return []
+        length = len(values[0])
+        if length == 0 or length > self._max_items:
+            return []
+        if any(len(slot) != length for slot in values):
+            return []
+        if any(type(value) is not str for slot in values for value in slot):
+            return []
+        wanted = {
+            index: tuple(text_digest(slot[index]) for slot in values)
+            for index in range(length)
+        }
+        receipts = self._receipts.get(run_key)
+        if not receipts:
+            return []
+        matches = [receipt for receipt in receipts if receipt.outputs() == wanted]
+        if matches:
+            self._receipts.move_to_end(run_key)
+        return matches
+
+    # -- the history capture (#403) -------------------------------------------
+
+    def recorded_fields(
+        self, prompt_id: Any, node_id: Any, values: Any
+    ) -> Optional[List[Fields]]:
+        """The history fields of the execution a cached node reuses, per list item.
+
+        ``values`` are the node's cached STRING outputs (``cached_string_outputs``).
+        Only for a node of the job currently executing, and only when every
+        receipt matching ``values`` kept the same fields; None otherwise. Never
+        counts and never marks the node replayed.
+        """
+        try:
+            with self._lock:
+                attempt = self._attempt
+                if (
+                    attempt is None
+                    or type(prompt_id) is not str
+                    or prompt_id != attempt.prompt_id
+                    or type(node_id) is not str
+                    or node_id not in attempt.nodes
+                ):
+                    return None
+                matches = self._matching(attempt.nodes[node_id][1], values)
+                if not matches:
+                    return None
+                found = []
+                for index in range(len(values[0])):
+                    fields = {receipt.items[index].fields for receipt in matches}
+                    if len(fields) != 1 or None in fields or _CONFLICT in fields:
+                        return None
+                    found.append(fields.pop())
+                return found
+        except Exception as error:
+            logger.debug(f"Could not read a cached run's history fields: {error}")
+            return None
+
 
 def _deduplicated(receipts: List[_Receipt]) -> List[_Receipt]:
-    """The newest receipt replaces older ones with the same outputs."""
+    """The newest receipt replaces older ones with the same outputs.
+
+    Counting needs only the outputs. When a replaced receipt recorded other
+    history fields for equal outputs (a raw text the outputs do not show),
+    the newest item's fields become a conflict, so no cached job of it is
+    given either text (#403). A conflict stays until the run key is dropped.
+    """
     if not receipts:
         return receipts
-    outputs = receipts[0].outputs()
-    return [receipts[0]] + [r for r in receipts[1:] if r.outputs() != outputs]
+    newest, outputs = receipts[0], receipts[0].outputs()
+    kept = [newest]
+    for receipt in receipts[1:]:
+        if receipt.outputs() != outputs:
+            kept.append(receipt)
+            continue
+        for index, item in list(newest.items.items()):
+            older = receipt.items[index].fields
+            if older == _CONFLICT or (
+                older is not None and item.fields is not None and older != item.fields
+            ):
+                newest.items[index] = item._replace(fields=_CONFLICT, field_bytes=0)
+    return kept
+
+
+def _bytes(receipts: List[_Receipt]) -> int:
+    """UTF-8 bytes of the history fields ``receipts`` keep."""
+    return sum(
+        item.field_bytes for receipt in receipts for item in receipt.items.values()
+    )
 
 
 def _job_nodes(graph: Dict[str, Any], outputs: Any) -> Dict[str, Tuple[str, str]]:

utils/execution_context_host.py

@@ -151,10 +151,77 @@ def cached_string_outputs(
         return None
 
 
+def _receipt_fields(
+    receipts: object, executor: object, job_id: str, source: object
+) -> list | None:
+    """Per list index, the ``(raw, prepend, append)`` of the execution reused (#403).
+
+    From the #347 receipt whose digests equal every STRING slot of the cache
+    entry; None when there is none or the answer is malformed.
+    """
+    try:
+        if receipts is None:
+            try:
+                from .cached_runs import CACHED_RUNS as receipts
+            except ImportError:
+                from utils.cached_runs import CACHED_RUNS as receipts
+        values = cached_string_outputs(executor, source.node_id, source.node_type)
+        if values is None:
+            return None
+        found = receipts.recorded_fields(job_id, source.node_id, values)
+        if type(found) is not list or len(found) != len(values[0]):
+            return None
+        position = (
+            (1 if source.output_slot in (1, 3) else 0)
+            if source.node_type == "PromptManagerPair"
+            else 0
+        )
+        fields = []
+        for slots in found:
+            slot = slots[position] if type(slots) is tuple else None
+            if not (
+                type(slot) is tuple
+                and len(slot) == 3
+                and all(type(value) is str for value in slot)
+            ):
+                return None
+            fields.append(slot)
+        return fields
+    except Exception:
+        return None
+
+
+def _agrees(inferred: PromptSnapshot, fields: tuple) -> bool:
+    """Every field the graph knows equals the receipt's; else fill nothing."""
+    known = (inferred.raw, inferred.prepend, inferred.append)
+    return all(
+        field.value is None or field.value == value
+        for field, value in zip(known, fields)
+    )
+
+
+def _filled(inferred: TextEvidence, value: str) -> TextEvidence:
+    """A field the graph cannot show, taken from the reused execution (#403)."""
+    if inferred.value is not None:
+        return inferred
+    return TextEvidence(value, "inferred", "graph_semantics", None)
+
+
 def observe_cached_context(
-    executor: object, message_type: object, message_data: object
+    executor: object,
+    message_type: object,
+    message_data: object,
+    *,
+    receipts: object = None,
 ) -> None:
-    """Copy bounded local strings without cache/provider execution or retention."""
+    """Copy bounded local strings without cache/provider execution or retention.
+
+    Raw, prepend and append come from graph inference; a field it leaves
+    unknown (runtime-only text) is taken from the receipt of the execution the
+    cache entry is proven to come from (``CachedRuns.recorded_fields``), so the
+    cached job is a tracked run of its prompt (#403). ``receipts`` defaults to
+    the process ``CACHED_RUNS``.
+    """
     try:
         binding = active_capture()
         if binding is None or type(message_type) is not str:
@@ -198,18 +265,29 @@ def observe_cached_context(
             inferred = manager.source_snapshot(cap, source)
             if inferred is None:
                 continue
+            fields = None
+            if any(
+                field.value is None
+                for field in (inferred.raw, inferred.prepend, inferred.append)
+            ):
+                fields = _receipt_fields(receipts, executor, cap.job_id, source)
             for index, value in enumerate(values):
                 if type(value) is not str:
                     continue
+                raw, prepend, append = inferred.raw, inferred.prepend, inferred.append
+                if fields is not None and _agrees(inferred, fields[index]):
+                    raw = _filled(raw, fields[index][0])
+                    prepend = _filled(prepend, fields[index][1])
+                    append = _filled(append, fields[index][2])
                 manager.observe(
                     cap,
                     source,
                     index,
                     PromptSnapshot(
-                        inferred.raw,
+                        raw,
                         TextEvidence(value, "recorded", "host_cache_output", None),
-                        inferred.prepend,
-                        inferred.append,
+                        prepend,
+                        append,
                     ),
                 )
     except Exception:

web/js/gallery-settings.js

@@ -40,7 +40,8 @@
         videoAutoplay: false,
         videoMute: true,
         videoLoop: true,
-        workspaceLayout: "grid",
+        // #451: Images opens in Masonry unless this browser chose another layout.
+        workspaceLayout: "masonry",
         workspaceDensity: "comfortable",
     });
 

web/js/workspace/collections.mjs

@@ -78,7 +78,8 @@ export function parseCollections(data) {
 }
 
 /** The sidebar's order while a re-read is pending: an approximation of the server's name key, then id. */
-const nameKey = name => name.normalize('NFC').toLowerCase().normalize('NFC');
+export const collectionNameKey = name => name.normalize('NFC').toLowerCase().normalize('NFC');
+const nameKey = collectionNameKey;
 const ordered = list => [...list].sort((a, b) => {
     const left = nameKey(a.name), right = nameKey(b.name);
     return left < right ? -1 : left > right ? 1 : a.id - b.id;

web/js/workspace/file-inspector.mjs

@@ -8,6 +8,7 @@ import { showWorkspaceToast } from './toast.mjs';
 import { openCopyFallback } from './workspace-dialogs.mjs';
 import { displayPreferences } from './preferences.mjs';
 import { fallBackToOriginal, thumbnailSource } from './thumbnail-support.mjs';
+import { openCollectionsDialog } from './inspector-collections.mjs';
 
 const RAW_KEYS = ['prompt', 'workflow', 'parameters', 'comfyassets_prompt_pairs'];
 const RAW_LIMIT = 'Raw metadata exceeds the response limit. Download the original file to inspect it externally.';
@@ -175,24 +176,32 @@ function extensionOf(record) {
     return (record.file?.extension || record.identity.path.split('.').at(-1)).replace(/^\./, '').toUpperCase();
 }
 /**
- * The chip row after the prompt blocks (D88): the file's current image tags first, then D21's linked tags not already
- * shown (case-insensitive). With `filterable` (a Tags filter exists, as in the Images panel) every image tag sets it, and
- * a linked tag only when it is the positive's own (D21); otherwise, as in the viewer, every chip is plain.
+ * The tag chips after the prompt blocks (D88, D154): the file's current image tags (`source: 'image'`), then D21's linked
+ * tags (`source: 'linked'`), each without case-insensitive repeats within its own source (D154: a linked tag that is also
+ * an image tag stays in both labelled sections). With `filterable` (a Tags filter exists, as in the Images panel) every
+ * image tag sets it, and a linked tag only when it is the positive's own (D21); otherwise, as in the viewer, every chip
+ * is plain.
  */
 export function inspectorTagChips(record, {filterable}) {
-    const shown = new Set(), chips = [];
-    const add = (names, filter) => {
+    const chips = [];
+    const add = (names, filter, source) => {
+        const shown = new Set();
         for (const name of names) {
             const key = name.toLowerCase();
             if (shown.has(key)) continue;
-            shown.add(key); chips.push({name, filter});
+            shown.add(key); chips.push({name, filter, source});
         }
     };
-    if (record.imageTags?.state === 'current') add(record.imageTags.tags, filterable);
+    if (record.imageTags?.state === 'current') add(record.imageTags.tags, filterable, 'image');
     const {library} = record;
-    if (library?.status === 'linked') add(library.tags, filterable && library.tags_scope === 'positive');
+    if (library?.status === 'linked') add(library.tags, filterable && library.tags_scope === 'positive', 'linked');
     return chips;
 }
+/** D154 (#449): a tag section's visible heading; linked tags name their owner as D146's category chip does. */
+export function tagSectionLabel(source, library) {
+    if (source === 'image') return 'Image tags';
+    return library?.tags_scope === 'pair' ? 'Recipe tags' : 'Prompt tags';
+}
 /** A concept chip whose visible text carries a visually hidden prefix naming what it is. */
 /**
  * The category chip (#332 D146): the hidden prefix, the category, then `span.chip-source` " · this image", " · prompt"
@@ -220,7 +229,12 @@ function prefixedChip(prefix, text) {
 /** D124: the inspector's line while "Load metadata on selection" is off. */
 export const DEFERRED_TEXT = 'Metadata loads when you ask for it.';
 
-export function mountFileInspector(container, {api, sender, parentScope, variant, inline = false, promptContext = '', onTagFilter, onRetry, onRefresh, onLoad, onInteraction = () => {}, copyFallback = openCopyFallback}) {
+/**
+ * #331 D139: with the client store `collections`, a listed file's footer ends with "Collections…", which opens its
+ * collections checklist; `onCollectionsChanged({added, removed})` hears the confirmed changes.
+ */
+export function mountFileInspector(container, {api, sender, parentScope, variant, inline = false, promptContext = '', onTagFilter, onRetry, onRefresh, onLoad, onInteraction = () => {}, copyFallback = openCopyFallback,
+    collections = null, onCollectionsChanged = null}) {
     if (parentScope.signal.aborted) return {setState() {}, dispose() {}};
     const scope = childScope(parentScope);
     let contentScope;
@@ -289,14 +303,11 @@ export function mountFileInspector(container, {api, sender, parentScope, variant
         // not what it matches, so they (and chips where no Tags filter exists, as in the viewer) stay plain. With the
         // D39 fallback a recipe without tags shows its positive's, which are filter buttons again. D88: the file's own
         // image tags lead the row and, where a Tags filter exists, always set it (it matches them too).
+        // D154 (#449): each source gets its own labelled section, so image tags never read as part of the negative prompt.
         const chips = inspectorTagChips(record, {filterable:!!onTagFilter});
-        if (chips.length) {
-            const tags = el('div', undefined, 'tags');
-            for (const {name, filter} of chips) {
-                if (!filter) {tags.append(el('span', name, 'chip')); continue;}
-                const chip = createButton(name, 'chip'); contentScope.listen(chip, 'click', () => onTagFilter(name)); tags.append(chip);
-            }
-            inner.append(tags);
+        for (const source of ['image', 'linked']) {
+            const own = chips.filter(chip => chip.source === source);
+            if (own.length) inner.append(tagSection(source, tagSectionLabel(source, library), own));
         }
         inner.append(...generationFacts(record, 'meta-grid', 'small'));
         if (full) inner.append(summaryButton(record), provenance(record, badge), rawMetadata(record));
@@ -317,8 +328,35 @@ export function mountFileInspector(container, {api, sender, parentScope, variant
             });
             footer.append(history);
         }
+        const member = collections ? revealTarget(record.identity) : null;
+        if (member) {
+            const button = createButton('');
+            button.dataset.collections = '';
+            button.append(createWorkspaceIcon('folder'), ' Collections…');
+            // A re-render replaces the button: closing focuses the current one, else the inspector's heading.
+            const fallback = () => container.querySelector('button[data-collections]') ?? container.querySelector('h2');
+            contentScope.listen(button, 'click', () => openCollectionsDialog({scope, collections, target:member, trigger:button,
+                onChanged:onCollectionsChanged, focusFallback:fallback}));
+            footer.append(button);
+        }
         if (!full && footer.childNodes.length) container.append(footer);
     }
+    /**
+     * D154: `section.tag-section` in the prompt sections' frame, its `.role-label` heading naming the source and count.
+     * The region's name leaves the count out and carries the prompt blocks' context ("Current file image tags").
+     */
+    function tagSection(source, label, chips) {
+        const section = el('section', undefined, `tag-section ${source}-tags`), heading = el('div', undefined, 'prompt-heading');
+        section.setAttribute('aria-label', promptContext ? `${promptContext} ${label.toLowerCase()}` : label);
+        heading.append(el('span', `${label} · ${chips.length}`, 'role-label'));
+        const tags = el('div', undefined, 'tags');
+        for (const {name: text, filter} of chips) {
+            if (!filter) {tags.append(el('span', text, 'chip')); continue;}
+            const chip = createButton(text, 'chip'); contentScope.listen(chip, 'click', () => onTagFilter(text)); tags.append(chip);
+        }
+        section.append(heading, tags);
+        return section;
+    }
     /**
      * D117 (LI-4): "Copy settings summary" after the six facts, the legacy COPY ALL. It copies as the concept's prompt
      * copy does: the shared toast, or the select-to-copy dialog where the clipboard is unavailable or refused.

web/js/workspace/images-category.mjs

@@ -4,7 +4,8 @@ import { countOf } from './plural.mjs';
  * The Images selection bar's "Set category" (#332, D140; the concept's bulk-category → simpleInput). Apply sets each
  * selected image's own category (`PUT /prompt_manager/images/category`); "Use prompt category" removes it after a
  * confirmation (`DELETE`), so the image shows its prompt's category again. It never writes a prompt: prompt
- * categories are set in the Prompt library editor and on the legacy admin page's "Set Category".
+ * categories are set in the Prompt library editor, the Prompt library's selection bar (#437, library-selection.mjs)
+ * and on the legacy admin page's "Set Category".
  */
 export const CATEGORY_URL = '/prompt_manager/images/category';
 export const IMAGE_CATEGORIES_URL = '/prompt_manager/images/categories';
@@ -20,11 +21,14 @@ export const CATEGORY_MAX_LENGTH = 100;
 // One request names at most 500 images (DC2), as "Add tags" does.
 export const MAX_CATEGORY_IMAGES = 500;
 
-/** The success toast of a set: unchanged files count as set; files no source lists are named after. */
-export function setOutcome(result, category) {
+/**
+ * The success toast of a set: unchanged items count as set; items that could not be found are named after. `noun` is
+ * what was set (the Prompt library's selection bar passes "prompt", #437).
+ */
+export function setOutcome(result, category, noun = 'image') {
     const updated = Number(result?.updated) || 0, unchanged = Number(result?.unchanged) || 0;
     const missing = Number(result?.not_found) || 0;
-    const done = `Category “${category}” set on ${countOf(updated + unchanged, 'image')}.`;
+    const done = `Category “${category}” set on ${countOf(updated + unchanged, noun)}.`;
     return missing ? `${done} ${missing.toLocaleString()} could not be found.` : done;
 }
 

web/js/workspace/images-collections.mjs

@@ -0,0 +1,76 @@
+/**
+ * The Images selection bar's collection actions (#331 D138, design §7.2–7.4):
+ *
+ * - **Add to collection** (the concept's `bulk-collection` simpleInput, app.js a1392–1398): "Add selected images to
+ *   collection", "Collection name", Add, with the existing names offered as suggestions. A name that matches a
+ *   collection ignoring case adds to it; a new name creates it. The selection stays and the listing is not reloaded
+ *   (an add never changes it).
+ * - **Remove from collection** (beyond the concept, only while a collection filter is active): no confirmation; the
+ *   listing then reloads through `onChanged`, keeping the checked files still listed.
+ *
+ * The routes take at most 500 explicit files. "All N matching" (#392, D135) is accepted only when every match is
+ * loaded and checked and there are at most 500 of them; otherwise the action toasts "Select at most 500 images." and
+ * sends nothing (ruling for #331 PR 3: the collection routes have no filter body).
+ */
+import { bulkTargets } from './images-tagging-bulk.mjs';
+import { ADD_DIALOG_ACTION, ADD_DIALOG_LABEL, ADD_DIALOG_TITLE, ADD_FAILED_TOAST, FILES_NOT_FOUND_TOAST, MAX_COLLECTION_NAME_LENGTH,
+    REMOVE_FAILED_TOAST, TOO_MANY_TOAST, addedMessage, isCollectionGone, normalizeCollectionName, removedMessage,
+    writeFailure } from './collections.mjs';
+import { openInputDialog } from './workspace-dialogs.mjs';
+import { showWorkspaceToast } from './toast.mjs';
+
+/**
+ * The files an action sends for the checked `images` (listing order): `{targets}`, or `{refusal}` with the toast.
+ * With `allMatching`, `total` is the listing's match count; only a fully loaded match is sent.
+ */
+export function collectionTargets(images, { allMatching = false, total = images.length } = {}) {
+    if (allMatching && total !== images.length) return { refusal: TOO_MANY_TOAST };
+    const targets = bulkTargets(images);
+    if (!targets) return { refusal: TOO_MANY_TOAST };
+    return targets.length ? { targets } : { refusal: FILES_NOT_FOUND_TOAST };
+}
+
+/**
+ * Add to collection over `images` (the checked files, taken now). `collections` is the client store; `openInput` is
+ * the dialog seam. A stale source (409) calls `onSourceChanged()` after its toast, so the page reloads the listing.
+ */
+export function openAddToCollection({ scope, collections, images, trigger = null, allMatching = false, total = images.length,
+    notify = showWorkspaceToast, openInput = openInputDialog, onSourceChanged = null }) {
+    const { targets, refusal } = collectionTargets(images, { allMatching, total });
+    if (refusal) { notify(refusal); return; }
+    openInput({ scope, trigger, title: ADD_DIALOG_TITLE, label: ADD_DIALOG_LABEL, action: ADD_DIALOG_ACTION,
+        maxLength: MAX_COLLECTION_NAME_LENGTH, suggestions: collections.collections.map(item => item.name),
+        onApply: async text => {
+            const name = normalizeCollectionName(text);
+            if (name === null) { notify(ADD_FAILED_TOAST); return; }
+            try {
+                const result = await collections.add({ name }, targets);
+                if (!scope.signal.aborted) notify(addedMessage(result));
+            } catch (error) {
+                if (scope.signal.aborted) return;
+                notify(writeFailure(error, ADD_FAILED_TOAST));
+                if (error?.code === 'source-changed') onSourceChanged?.();
+            }
+        } });
+}
+
+/**
+ * Remove the checked `images` from the open collection `collectionId`. Resolves once the toast is shown; a confirmed
+ * remove, or a collection found deleted meanwhile, calls `onChanged()` so the page lists it again.
+ */
+export async function removeFromCollection({ scope, collections, collectionId, images, allMatching = false, total = images.length,
+    notify = showWorkspaceToast, onChanged = null, onSourceChanged = null }) {
+    const { targets, refusal } = collectionTargets(images, { allMatching, total });
+    if (refusal) { notify(refusal); return; }
+    try {
+        const result = await collections.remove(collectionId, targets);
+        if (scope.signal.aborted) return;
+        notify(removedMessage(result.collection.name, result.removed));
+        onChanged?.();
+    } catch (error) {
+        if (scope.signal.aborted) return;
+        notify(writeFailure(error, REMOVE_FAILED_TOAST));
+        if (isCollectionGone(error)) onChanged?.();
+        else if (error?.code === 'source-changed') onSourceChanged?.();
+    }
+}

web/js/workspace/images-selection.mjs

@@ -8,6 +8,7 @@ import { closeWorkspaceDialog, openContentDialog, openInputDialog } from './work
 import { TOO_MANY, bulkTargets, openAddTags, openAutoTag } from './images-tagging-bulk.mjs';
 import { showWorkspaceToast } from './toast.mjs';
 import { applyKeys, createLongPress, matchingLabel, matchingOffer, offersAllMatching, rangeKeys } from './images-multiselect.mjs';
+import { openAddToCollection, removeFromCollection } from './images-collections.mjs';
 
 const CHECKBOX = '.image-overlays input[type="checkbox"]';
 const CHECKBOX_LABEL = '.image-overlays > label';
@@ -42,9 +43,14 @@ function el(tag, text, className) {
  * `clear` serve the grid's keys (images.mjs). `getListing()` gives `{total, hasMore, positiveCount}` and
  * `getFilter()` the listing filter: once every loaded image is checked and more match, the bar offers "Select all N
  * matching", whose actions send that filter to the server.
+ *
+ * Collections (#331 D138, design §7): with the client store `collections`, Add to collection follows Set category;
+ * while a collection filter is active (`collection`, its id) Remove from collection follows it, and a confirmed remove
+ * calls `onCollectionsChanged()` to reload the listing. A stale source on either calls `onSourceChanged()`.
  */
 export function mountSelectionBar({ scope, api, grid, getItems, focusFallback, onCategorySet, activity = null, onTagged,
-    getListing, getFilter, getCategories, checked: initial = Object.freeze(new Set()) }) {
+    getListing, getFilter, getCategories, checked: initial = Object.freeze(new Set()), collections = null,
+    collection = null, onCollectionsChanged = null, onSourceChanged = null }) {
     let checked = initial;
     let exporting = false, categorizing = false;
     // D135: the range anchor, "all N matching" mode, touch selection mode and a pending Shift+click on a checkbox.
@@ -75,7 +81,19 @@ export function mountSelectionBar({ scope, api, grid, getItems, focusFallback, o
             ...matchingTarget() }));
         bulk.append(addTags, autoTag);
     }
-    bulk.append(setCategory, exportMetadata, clearButton);
+    bulk.append(setCategory);
+    if (collections) {
+        const addToCollection = createButton('Add to collection');
+        scope.listen(addToCollection, 'click', () => openAddToCollection({ scope, collections, images: checkedImages(),
+            trigger: addToCollection, ...collectionSelection(), onSourceChanged }));
+        bulk.append(addToCollection);
+    }
+    if (collections && collection) {
+        const removeButton = createButton('Remove from collection');
+        scope.listen(removeButton, 'click', () => removeChecked(removeButton));
+        bulk.append(removeButton);
+    }
+    bulk.append(exportMetadata, clearButton);
 
     /** The count updates in place: the same bar and `strong`, its text written only when it changes. */
     function render() {
@@ -143,6 +161,24 @@ export function mountSelectionBar({ scope, api, grid, getItems, focusFallback, o
     function matchingTarget() {
         return allMatching ? { filter: getFilter(), count: getListing().total } : {};
     }
+    /** The collection actions' view of "all N matching" (#331 ruling): the loaded images and the match count. */
+    function collectionSelection() {
+        return { allMatching, total: allMatching ? getListing().total : checked.size };
+    }
+    /** Remove from collection: one write at a time; the selection stays (a reload keeps what is still listed). */
+    let removing = false;
+    async function removeChecked(button) {
+        if (removing) return;
+        removing = true;
+        const release = busy(button);
+        try {
+            await removeFromCollection({ scope, collections, collectionId: Number(collection), images: checkedImages(),
+                ...collectionSelection(), onChanged: onCollectionsChanged, onSourceChanged });
+        } finally {
+            removing = false;
+            release();
+        }
+    }
     /** Toggle one image on its own; it becomes the anchor. */
     function toggle(key, on = !isChecked(key)) {
         settleMatching();

web/js/workspace/images-state.mjs

@@ -1,23 +1,39 @@
 import { noun } from './plural.mjs';
 
 /** Identity is the source and relative path; content hashes may be shared by files. */
+/** The layout and size a saved view, an embedded gallery and a page without readable preferences use. */
+const FIXED_VIEW = Object.freeze({ layout: 'grid', density: 'comfortable' });
+
 /**
- * The layout and thumbnail size an Images visit starts with (#249 System S3). The last visit (the current view) keeps
- * its own; a saved view starts as before, at Grid and Comfortable; any other new visit starts from System ›
- * Appearance's defaults (`preferences`, preferences.mjs values, or null when unavailable).
+ * The defaults a new Images visit starts from: System › Appearance's (`preferences`, preferences.mjs values; Masonry
+ * when the layout is missing, #451), or the fixed Grid/Comfortable when they are unavailable (null).
+ */
+export function visitDefaults(preferences) {
+    if (!preferences) return { ...FIXED_VIEW };
+    return { layout: preferences.workspaceLayout || 'masonry', density: preferences.workspaceDensity || 'comfortable' };
+}
+
+/**
+ * The layout and thumbnail size an Images visit starts with (#249 System S3). A saved view starts as before, at Grid
+ * and Comfortable; any other new visit starts from `visitDefaults(preferences)`. The last visit (`prior`, the current
+ * view) keeps a value its toolbar changed; a value it kept from the defaults it started from (`prior.defaults`) follows
+ * defaults saved since, in this tab or another (#451). A prior without `defaults` keeps its own values.
  */
 export function initialImagesView({ prior = null, saved = null, preferences = null } = {}) {
-    if (prior) return { layout: prior.layout || 'grid', density: prior.density || 'comfortable' };
-    if (saved || !preferences) return { layout: 'grid', density: 'comfortable' };
-    return { layout: preferences.workspaceLayout || 'grid', density: preferences.workspaceDensity || 'comfortable' };
+    if (!prior) return saved ? { ...FIXED_VIEW } : visitDefaults(preferences);
+    const own = { layout: prior.layout || FIXED_VIEW.layout, density: prior.density || FIXED_VIEW.density };
+    if (!prior.defaults || !preferences) return own;
+    const current = visitDefaults(preferences);
+    const follow = key => own[key] === prior.defaults[key] ? current[key] : own[key];
+    return { layout: follow('layout'), density: follow('density') };
 }
 
 export function imageKey(image) {
     return `${image.root_index}:${image.path}`;
 }
 
-/** The workspace listing's sorts: the concept's Recently added, Name A–Z and Highest rated, plus Oldest and Z–A (D14). */
-export const IMAGE_SORTS = Object.freeze(['newest', 'oldest', 'name', 'name-desc', 'rating']);
+/** The workspace listing's sorts: the concept's Recently added and Name A–Z, plus Oldest and Z–A (D14; no rating, D153). */
+export const IMAGE_SORTS = Object.freeze(['newest', 'oldest', 'name', 'name-desc']);
 
 const isCount = value => Number.isSafeInteger(value) && value >= 0;
 // The largest page a refresh asks for (the listing clamps at 500).
@@ -299,7 +315,7 @@ const startOfDay = date => new Date(date.getFullYear(), date.getMonth(), date.ge
 
 /**
  * D17: the caption's right-hand date, "<Today|Yesterday|weekday> · <Month D>" of the newest result under Newest first,
- * with ", YYYY" outside the current year; '' under Oldest first and the name or rating sorts.
+ * with ", YYYY" outside the current year; '' under Oldest first and the name sorts.
  */
 export function captionDate(items, sort, now = new Date()) {
     if (sort !== 'newest') return '';

web/js/workspace/images.mjs

@@ -1,5 +1,5 @@
 import { createScope, createButton } from '../shared/toolkit/core.js';
-import { createImageStream, withItem, imageKey, countLine, captionDate, scopeLabel, activeRefinements, initialImagesView, listingFilter } from './images-state.mjs';
+import { createImageStream, withItem, imageKey, countLine, captionDate, scopeLabel, activeRefinements, initialImagesView, visitDefaults, listingFilter } from './images-state.mjs';
 import { createImageCard } from './image-card.mjs';
 import { layoutImages, visibleImages, gridMetrics, stacksCardInfo } from './image-layout.mjs';
 import { mountFileInspector } from './file-inspector.mjs';
@@ -61,8 +61,9 @@ function selectOf(id, options, value) {
     return select;
 }
 
-// The concept's three sorts first, then the production ones it lacks (D14); values are the listing's sorts (B3).
-const IMAGE_SORT_OPTIONS = Object.freeze([['newest', 'Recently added'], ['name', 'Name A–Z'], ['rating', 'Highest rated'],
+// The concept's sorts first, then the production ones it lacks (D14); values are the listing's sorts (B3). D153 (#447)
+// drops the concept's "Highest rated".
+const IMAGE_SORT_OPTIONS = Object.freeze([['newest', 'Recently added'], ['name', 'Name A–Z'],
     ['oldest', 'Oldest first'], ['name-desc', 'Name Z–A']]);
 const MEDIA_OPTIONS = Object.freeze([['all', 'All media'], ['image', 'Images'], ['video', 'Videos']]);
 const RETRY_RESCAN = 'Rescan from Import to retry.';
@@ -116,10 +117,12 @@ export function selectedImage() {
  * Images last visit. `collection` (an id) is a collection page (#331 D137): only its members are listed, favorites is
  * off, and it neither restores nor replaces the last visit either. A listing that answers `collection-not-found` calls
  * `onCollectionMissing()` when given (the shell shows its deleted-collection page); otherwise (a saved view on a
- * deleted collection, D140) the page lists nothing and says so.
+ * deleted collection, D140) the page lists nothing and says so. `collections` (the shell's client store) adds the
+ * selection bar's Add to collection, and Remove from collection while a collection is open (D138).
  */
 export function mountImages(main, api, parentScope, {embedded = false, fetchPage, emptyText, focusTarget = null, savedState = null,
-    favorites = false, collection: collectionId = null, onCollectionMissing = null, onSaveView = () => {}, activity = null} = {}) {
+    favorites = false, collection: collectionId = null, onCollectionMissing = null, onSaveView = () => {}, activity = null,
+    collections = null} = {}) {
     const scope = createScope();
     const dispose = () => scope.dispose();
     parentScope.signal.addEventListener('abort', dispose, {once:true});
@@ -154,6 +157,8 @@ export function mountImages(main, api, parentScope, {embedded = false, fetchPage
     // The layout and size defaults are the Images page's: an embedded gallery (a prompt history's, its dialogs) has no
     // toolbar to change them, so it keeps its fixed Grid/Comfortable. The metadata and card defaults apply everywhere.
     const start = initialImagesView({prior, saved, preferences:embedded ? null : preferences});
+    // The defaults this visit started from: a return follows defaults saved since unless the toolbar changed a value (#451).
+    const startDefaults = visitDefaults(embedded ? null : preferences);
     let density = start.density;
     let metrics = {...INITIAL_METRICS};
     let layout = start.layout, selection = visit?.selection || null, frame = 0, loadFrame = 0, restoreFrame = 0, observer;
@@ -180,7 +185,7 @@ export function mountImages(main, api, parentScope, {embedded = false, fetchPage
         // A Set category write still in flight may change what the saved collection shows: the return reloads instead.
         // A favorites-only collection or a saved view is not the plain Images collection, so it never becomes the
         // Images return state (#249 D51, and M2 for saved views).
-        if (!embedded && !filters.favorites && !filters.collection && !saved) lastVisit = state && !bar.writing() ? { filters: {...filters, tags: [...filters.tags]}, state, layout, density, selection, checked: bar.checked(), scrollY: window.scrollY } : null;
+        if (!embedded && !filters.favorites && !filters.collection && !saved) lastVisit = state && !bar.writing() ? { filters: {...filters, tags: [...filters.tags]}, state, layout, density, defaults: startDefaults, selection, checked: bar.checked(), scrollY: window.scrollY } : null;
         cancelAnimationFrame(restoreFrame);
     });
     main.classList.add('images-main');
@@ -302,10 +307,20 @@ export function mountImages(main, api, parentScope, {embedded = false, fetchPage
         getFilter:() => {
             const filter = listingFilter(filters), scope = state?.items[0]?.root_scope;
             return filter.root !== undefined && typeof scope === 'string' ? {...filter, root_scope:scope} : filter;
-        }});
+        },
+        // #331 D138: Add to collection, and Remove from collection on an open collection, which then lists it again.
+        collections, collection:filters.collection || null, onCollectionsChanged:collectionChanged,
+        onSourceChanged:refreshInspectionCollection});
     if (bar) main.append(bar.element);
     main.append(body);
+    // #331 D139: a change the inspector or the viewer confirmed that touches the open collection lists it again; the
+    // viewer reports only once it closes.
+    const inspectorCollections = embedded ? null : collections;
+    const collectionsChangedBy = ({added, removed}) => {
+        if (filters.collection && [...added, ...removed].includes(Number(filters.collection))) collectionChanged();
+    };
     const mediaViewer = createMediaViewer({api,container:main,parentScope:scope,inspectionController:inspection,sender,
+        collections:inspectorCollections,onCollectionsChanged:collectionsChangedBy,
         onSelect:image => {if (selection !== imageKey(image)) inspect(image);else inspection.select(image);},
         onRefresh:refreshInspectionCollection,onClose:scheduleLoad,onMediaError:image => recovery.recover(image),onRestoreFocus:image => {
             const index = state?.items.findIndex(entry => imageKey(entry) === imageKey(image));
@@ -322,7 +337,8 @@ export function mountImages(main, api, parentScope, {embedded = false, fetchPage
         selectedImage = image;
         for (const video of details.querySelectorAll('video')) releaseVideo(video);
         detailInspector = mountFileInspector(details, {sender, parentScope:scope, variant:'panel',
-            onTagFilter:embedded ? undefined : filterByTag, onRetry:() => inspection.retry(), onLoad:() => inspection.load(), onRefresh:refreshInspectionCollection});
+            onTagFilter:embedded ? undefined : filterByTag, onRetry:() => inspection.retry(), onLoad:() => inspection.load(), onRefresh:refreshInspectionCollection,
+            collections:inspectorCollections, onCollectionsChanged:collectionsChangedBy});
         detailInspector.setState(inspection.state);
         inspection.select(image);
         for (const [key, card] of rendered) markSelected(card, key === selection);
@@ -734,6 +750,19 @@ export function mountImages(main, api, parentScope, {embedded = false, fetchPage
         const band = parseFloat(getComputedStyle(optionsStatus).scrollMarginBlockEnd) || 0;
         if (bottom > innerHeight - band && top < innerHeight) optionsStatus.scrollIntoView({block:'nearest', inline:'nearest'});
     }
+    /**
+     * A confirmed collection change that may change this listing (a remove from the open collection): the loaded range
+     * is read again in place, keeping the scroll and the checked files still listed. A failed re-read lists the page
+     * again from its start, whose failure shows the usual Retry (or the deleted-collection page).
+     */
+    async function collectionChanged() {
+        if (scope.signal.aborted || !filters.collection || recovering) return;
+        try {
+            await stream.refresh();
+        } catch {
+            if (!scope.signal.aborted) reset();
+        }
+    }
     async function refreshInspectionCollection() {
         if (scope.signal.aborted) return;
         recovery.reset();

web/js/workspace/index.mjs

@@ -9,7 +9,7 @@ import { mountVocabulary } from './vocabulary.mjs';
 import { createWorkspaceIcon } from './icons.mjs';
 import { bindCanvasLink, canvasHref } from './topbar.mjs';
 import { renderHeading, pageHeading } from './heading.mjs';
-import { closeWorkspaceDialog } from './workspace-dialogs.mjs';
+import { closeReviewDialog, closeWorkspaceDialog } from './workspace-dialogs.mjs';
 import { createSavedViews, openSaveViewDialog, openEditViewDialog, savedViewHash } from './saved-views.mjs';
 import { createCollections, collectionHash, openEditCollectionDialog, missingNote, DELETED_PAGE_TEXT, OPEN_IMAGES_LINK,
     LIST_FAILED_TEXT } from './collections.mjs';
@@ -20,7 +20,7 @@ import { mountInspect } from './inspect.mjs';
 import { mountReview } from './review.mjs';
 import { createPromptSender } from './prompt-sender.mjs';
 import { mountSystem } from './system.mjs';
-import { onRestoreFinished, onRestoreHandedBack, showRestoreReview } from './system-backups.mjs';
+import { createRestoreReload, onRestoreFinished, onRestoreHandedBack, showRestoreReview } from './system-backups.mjs';
 import { mountLibraryBanner } from './library-location.mjs';
 import { detectJxlSupport } from './thumbnail-support.mjs';
 
@@ -184,13 +184,20 @@ function mountWorkspace() {
     refreshCounts(NAV_COUNT_ROUTES);
     // S9: a finished restore replaced the library. The cached visits are dropped, the counts and saved views read
     // again, and the open page is shown again from the restored data; the Activity job view the restore opened stays.
-    scope.cleanup(onRestoreFinished(activity, () => {
+    // #415: an open review closes (its plan no longer describes storage); while any other dialog is open, all of it
+    // waits until the last one has closed, so the dialog keeps what was typed.
+    const restoreReload = createRestoreReload(() => {
         forgetImagesVisit();
         forgetLibraryVisit();
         refreshCounts(NAV_COUNT_ROUTES);
         savedViews.load();
         collections.load();
         if (currentKey !== 'activity') navigateTo(window.location.hash);
+    });
+    scope.cleanup(() => restoreReload.dispose());
+    scope.cleanup(onRestoreFinished(activity, () => {
+        closeReviewDialog();
+        restoreReload.request();
     }));
     // S9 I1: a restore this page started that could not run yet (other requests kept reading) hands its review back:
     // it opens again over the page shown now, with Restore (again) and Discard; leaving it any other way discards it.
@@ -403,7 +410,7 @@ function mountWorkspace() {
         if (route.id === 'images' || route.id === 'favorites' || route.id === 'collection') {
             mountNewPromptAction(main, api, view);
             images = mountImages(main, api, view, {savedState:saved?.state, favorites:route.id === 'favorites',
-                collection, onCollectionMissing, onSaveView:saveView('images', view), activity});
+                collection, onCollectionMissing, onSaveView:saveView('images', view), activity, collections});
             return;
         }
         if (route.id === 'library') {

web/js/workspace/inspector-collections.mjs

@@ -0,0 +1,189 @@
+/**
+ * The shared inspector's "Collections…" dialog (#331 D139, design §8), beyond the concept: one listed file's
+ * collections as a checklist in the sidebar's order, a New collection field, then Cancel and Save.
+ *
+ * - The file's membership is read when the dialog opens ("Loading collections…"); a failed read says "Collections
+ *   could not be loaded." with Retry, and Save waits for a read that worked.
+ * - Each row is the concept's checkbox row (`label.tag-suggestion`, app.js a1128). With no collection the list says
+ *   "No collections yet. Name one below."
+ * - Save sends one request per change, in order: the new name, each newly checked collection, each unchecked one.
+ *   Nothing changed sends nothing and closes. All sent: the dialog closes with "Collections updated.". A failure keeps
+ *   it open with "The collections could not be updated. Try again." and the membership read again; the changes that
+ *   went through still reach `onChanged({added, removed})` (collection ids), as a full success does, but only once the
+ *   dialog closes: a page that lists the file again (a collection page) would otherwise remove the dialog before its
+ *   failure is read. Save waits for the membership read again.
+ * - A New collection name equal to a listed collection ignoring case is that collection's checkbox, checked. A name
+ *   that cannot be a collection name is refused on the field; nothing is sent and the checkbox edits stay.
+ * - Focus starts on the field and moves to the first checkbox once they are listed; closing returns it to the
+ *   trigger, or to `focusFallback()` when an inspector re-render removed the trigger.
+ */
+import { createButton } from '../shared/toolkit/core.js';
+import { closeWorkspaceDialog, field, openContentDialog } from './workspace-dialogs.mjs';
+import { showWorkspaceToast } from './toast.mjs';
+import { COLLECTIONS_UPDATED_TOAST, MAX_COLLECTION_NAME_LENGTH, MEMBERSHIP_FAILED_TEXT, MEMBERSHIP_LOADING_TEXT,
+    NO_COLLECTIONS_TEXT, UPDATE_FAILED_TEXT, collectionNameKey, normalizeCollectionName } from './collections.mjs';
+
+export const COLLECTIONS_DIALOG_TITLE = 'Collections';
+export const NEW_COLLECTION_LABEL = 'New collection';
+
+function el(tag, text, className) {
+    const node = document.createElement(tag);
+    if (text !== undefined) node.textContent = text;
+    if (className) node.className = className;
+    return node;
+}
+
+/**
+ * Open the dialog for `target` (`{root_index, root_scope, path}`) with the client store `collections`. `openDialog`
+ * and `closeDialog` are the dialog seams; `notify` the toast.
+ */
+export function openCollectionsDialog({ scope, collections, target, trigger = null, notify = showWorkspaceToast, onChanged = null,
+    focusFallback = null, openDialog = openContentDialog, closeDialog = closeWorkspaceDialog }) {
+    const local = new AbortController();
+    const active = () => !scope.signal.aborted && !local.signal.aborted;
+    let member = null, saving = false, listed = false;
+    // The confirmed changes, reported once when the dialog closes (review M2).
+    const changes = { added: [], removed: [] };
+    const status = el('p', '', 'small');
+    const retry = createButton('Retry');
+    const list = el('div');
+    list.setAttribute('role', 'group');
+    list.setAttribute('aria-label', COLLECTIONS_DIALOG_TITLE);
+    const input = el('input');
+    input.value = '';
+    input.maxLength = MAX_COLLECTION_NAME_LENGTH;
+    input.autocomplete = 'off';
+    input.addEventListener('input', () => input.setCustomValidity?.(''));
+    const problem = el('p', '', 'small');
+    problem.setAttribute('role', 'alert');
+    problem.hidden = true;
+    const actions = el('div', undefined, 'dialog-actions');
+    const cancel = createButton('Cancel');
+    const save = createButton('Save', 'ca-primary');
+    actions.append(cancel, save);
+    const blocked = () => member === null || saving;
+    const paintSave = () => {
+        if (blocked()) save.setAttribute('aria-disabled', 'true');
+        else save.removeAttribute('aria-disabled');
+    };
+    retry.addEventListener('click', () => load());
+    cancel.addEventListener('click', () => { if (!saving) closeDialog(); });
+    save.addEventListener('click', () => { if (!blocked()) submit(); });
+
+    /** One checkbox row per collection, checked where the file is a member. */
+    function renderList() {
+        list.replaceChildren();
+        if (!collections.collections.length) {
+            list.append(el('p', NO_COLLECTIONS_TEXT, 'small'));
+            return;
+        }
+        for (const item of collections.collections) {
+            const row = el('label', undefined, 'tag-suggestion');
+            const box = el('input');
+            Object.assign(box, { type: 'checkbox', id: `collection-choice-${item.id}`, checked: member.has(item.id) });
+            box.dataset.collectionId = String(item.id);
+            row.htmlFor = box.id;
+            row.append(box, el('span', item.name));
+            list.append(row);
+        }
+    }
+    function showStatus(text, failed) {
+        status.textContent = text;
+        status.setAttribute('role', failed ? 'alert' : 'status');
+        status.hidden = !text;
+        retry.hidden = !failed;
+    }
+    /** Save waits while the membership is read (again). */
+    function loading() {
+        member = null;
+        paintSave();
+        list.replaceChildren();
+        showStatus(MEMBERSHIP_LOADING_TEXT, false);
+    }
+    async function load() {
+        loading();
+        try {
+            const ids = await collections.membership(target);
+            if (!active()) return;
+            member = new Set(ids);
+            showStatus('', false);
+            renderList();
+            // The first listing takes focus from the untouched field to the first checkbox (review L6).
+            if (!listed && !input.value) firstBox().focus();
+            listed = true;
+        } catch {
+            if (!active()) return;
+            showStatus(MEMBERSHIP_FAILED_TEXT, true);
+        }
+        paintSave();
+    }
+    const boxesOf = () => [...list.children].map(row => row.children?.[0]).filter(box => box?.dataset?.collectionId);
+    const firstBox = () => boxesOf()[0] ?? input;
+    /**
+     * The requests Save sends, in order; null when the new name cannot be a collection name. A name equal to a listed
+     * collection's ignoring case is that collection's checkbox, checked (review L4).
+     */
+    function plan() {
+        const typed = input.value.trim();
+        const name = typed ? normalizeCollectionName(typed) : '';
+        if (name === null) return null;
+        const named = name ? collections.collections.find(item => collectionNameKey(item.name) === collectionNameKey(name)) : null;
+        const steps = name && !named ? [{ add: { name } }] : [];
+        const boxes = boxesOf().map(box => ({ id: Number(box.dataset.collectionId), checked: box.checked }))
+            .map(box => (box.id === named?.id ? { ...box, checked: true } : box));
+        for (const box of boxes) if (box.checked && !member.has(box.id)) steps.push({ add: { collectionId: box.id } });
+        for (const box of boxes) if (!box.checked && member.has(box.id)) steps.push({ remove: box.id });
+        return steps;
+    }
+    async function submit() {
+        problem.textContent = '';
+        problem.hidden = true;
+        const steps = plan();
+        if (steps === null) {
+            // Review L5: refused on the field; nothing is sent and the checkbox edits stay.
+            input.setCustomValidity?.(UPDATE_FAILED_TEXT);
+            input.reportValidity?.();
+            input.focus();
+            return;
+        }
+        if (steps.length === 0) { closeDialog(); return; }
+        saving = true;
+        paintSave();
+        let failed = false;
+        for (const step of steps) {
+            try {
+                if (step.add) changes.added.push((await collections.add(step.add, [target])).collection.id);
+                else changes.removed.push((await collections.remove(step.remove, [target])).collection.id);
+            } catch {
+                failed = true;
+                break;
+            }
+            if (!active()) return;
+            if (step.add?.name) input.value = '';
+        }
+        if (!active()) return;
+        saving = false;
+        if (!failed) {
+            closeDialog();
+            notify(COLLECTIONS_UPDATED_TOAST);
+            return;
+        }
+        problem.textContent = UPDATE_FAILED_TEXT;
+        problem.hidden = false;
+        // Review M1: Save waits from here until the membership is read again. A collection deleted meanwhile is
+        // re-read by the store; the list then shows what the server holds.
+        loading();
+        await collections.settled();
+        if (active()) await load();
+    }
+    function closed() {
+        local.abort();
+        if (changes.added.length || changes.removed.length) onChanged?.({ added: [...changes.added], removed: [...changes.removed] });
+        // The shell focuses a trigger still on the page; one an inspector re-render replaced falls back (review L6).
+        if (trigger && !trigger.isConnected) focusFallback?.()?.focus();
+    }
+    showStatus('', false);
+    openDialog({ scope, title: COLLECTIONS_DIALOG_TITLE, trigger, canClose: () => !saving, onClose: closed, focusTarget: input,
+        nodes: [status, retry, list, field(NEW_COLLECTION_LABEL, input, 'collection-new'), problem, actions] });
+    load();
+}

web/js/workspace/library-editor-state.mjs

@@ -9,7 +9,9 @@
  */
 const MODES = new Set(['new', 'metadata', 'revision', 'negative']);
 const KINDS = new Set(['metadata', 'create', 'add-negative', 'archive-pair', 'restore-pair']);
-const FIELDS = ['title', 'positive_text', 'negative_text', 'category', 'tags', 'notes', 'rating'];
+// D153 (#447): the workspace edits no 1–5 rating; a stored one is left exactly as it is.
+export const EDITOR_FIELDS = Object.freeze(['title', 'positive_text', 'negative_text', 'category', 'tags', 'notes']);
+const FIELDS = EDITOR_FIELDS;
 const RECOVERY = {'metadata-conflict': 'conflict', 'storage-changed': 'storage-changed'};
 // A proxy timeout or bad gateway cannot tell whether the server committed the write.
 const UNCONFIRMED_STATUS = new Set([502, 503, 504]);
@@ -106,27 +108,20 @@ export function fallbackTitle(text) {return text.split('\n')[0].trim();}
 export function parseTags(value) {return [...new Set(value.split(',').map(tag => tag.trim()).filter(Boolean))];}
 const sameTags = (a, b) => a.length === b.length && new Set([...a, ...b]).size === a.length;
 const stored = value => (value ?? '').trim();
-function normalizeRating(value) {
-    if (value === null || value === '') return null;
-    const number = typeof value === 'string' && /^[1-5]$/.test(value) ? Number(value) : value;
-    if (!rating(number)) throw new TypeError('Rating must be 1 to 5 or null');
-    return number;
-}
-
-const blankDraft = () => ({title: '', positive_text: '', negative_text: '', category: '', tags: '', notes: '', rating: null});
+const blankDraft = () => ({title: '', positive_text: '', negative_text: '', category: '', tags: '', notes: ''});
 function seedDraft(mode, {owner, pair}) {
     if (!owner) return blankDraft();
     if (mode === 'negative') return {...blankDraft(), positive_text: owner.text};
     return {title: stored(owner.title) ? owner.title : fallbackTitle(owner.text), positive_text: owner.text,
         negative_text: pair ? pair.negative_text : '', category: owner.category ?? '', tags: owner.tags.join(', '),
-        notes: owner.notes ?? '', rating: owner.rating};
+        notes: owner.notes ?? ''};
 }
 /** Keep fields the user changed from the previous seed; the rest follow the new seed. */
 function overlay(next, draft, previous) {
     return Object.fromEntries(FIELDS.map(name => [name, draft[name] !== previous[name] ? draft[name] : next[name]]));
 }
 function baselineOf(owner) {
-    return owner && {title: owner.title, category: owner.category, tags: [...owner.tags], rating: owner.rating, notes: owner.notes};
+    return owner && {title: owner.title, category: owner.category, tags: [...owner.tags], notes: owner.notes};
 }
 
 /** Only changed fields; an untouched fallback title is not a title edit. */
@@ -136,19 +131,18 @@ function metadataChanges(draft, seed, baseline) {
     if (draft.title !== seed.title && title !== stored(baseline.title)) changes.title = title;
     if (category !== stored(baseline.category)) changes.category = category;
     if (!sameTags(tags, baseline.tags)) changes.tags = tags;
-    if (draft.rating !== baseline.rating) changes.rating = draft.rating;
     if (draft.notes !== (baseline.notes ?? '')) changes.notes = draft.notes;
     return changes;
 }
 function createMetadata(draft) {
     return {title: draft.title.trim(), category: draft.category.trim() || UNCATEGORIZED, tags: parseTags(draft.tags),
-        rating: draft.rating, notes: draft.notes};
+        notes: draft.notes};
 }
 // A blank category is stored as NULL; "Uncategorized" is only the UI's label for it.
 const categoryKey = value => (stored(value) === UNCATEGORIZED ? '' : stored(value));
 function sameMetadata(sent, owner) {
     return sent.title === stored(owner.title) && categoryKey(sent.category) === categoryKey(owner.category) && sameTags(sent.tags, owner.tags) &&
-        sent.rating === owner.rating && sent.notes === (owner.notes ?? '');
+        sent.notes === (owner.notes ?? '');
 }
 function draftProblem(mode, draft) {
     if (mode !== 'negative' && !draft.title.trim()) return Object.assign(new Error('Title is required'), {code: 'invalid-draft'});
@@ -328,9 +322,8 @@ export function createLibraryEditorController({api, mode, positiveId = null, pai
         },
         setField(name, value) {
             if (!FIELDS.includes(name)) throw new TypeError(`Unknown editor field: ${name}`);
-            const next = name === 'rating' ? normalizeRating(value) : value;
-            if (name !== 'rating' && typeof next !== 'string') throw new TypeError(`${name} must be a string`);
-            emit({draft: {...state.draft, [name]: next}});
+            if (typeof value !== 'string') throw new TypeError(`${name} must be a string`);
+            emit({draft: {...state.draft, [name]: value}});
         },
         async submit() {
             if (disposed || !SUBMITTABLE.has(state.phase)) return;

web/js/workspace/library-editor.mjs

@@ -7,12 +7,11 @@
  * module only renders `controller.state` and forwards user actions.
  */
 import { createButton } from '../shared/toolkit/core.js';
-import { createLibraryEditorController } from './library-editor-state.mjs';
+import { createLibraryEditorController, EDITOR_FIELDS } from './library-editor-state.mjs';
 
 const TITLES = {new: 'New prompt', metadata: 'Edit recipe metadata', revision: 'Create text revision', negative: 'Add a negative variant'};
-const RATINGS = [['5', '5 / 5'], ['4', '4 / 5'], ['3', '3 / 5'], ['2', '2 / 5'], ['1', '1 / 5'], ['0', 'Unrated']];
 const FIELD_IDS = {title: 'recipe-title', positive_text: 'recipe-positive', negative_text: 'recipe-negative', category: 'recipe-category',
-    tags: 'recipe-tags', rating: 'recipe-rating', notes: 'recipe-notes'};
+    tags: 'recipe-tags', notes: 'recipe-notes'};
 const SUBMITTABLE = new Set(['ready', 'write-error']);
 const NEGATIVE_HELP = 'Leave empty for an explicitly empty negative. The positive stays unchanged.';
 const NO_PAIR = 'Archive needs a selected saved negative variant.';
@@ -218,9 +217,8 @@ export function createLibraryEditor({api, parentScope, onCommitted = () => {}, o
             category.setAttribute('list', 'categories');
             tagInput.setAttribute('list', 'tag-options');
             columns.append(field('Category', category, FIELD_IDS.category), field('Tags, separated by commas', tagInput, FIELD_IDS.tags));
-            const rating = controls.rating = el('select');
-            for (const [value, text] of RATINGS) rating.add(new Option(text, value));
-            form.append(columns, categories, tags, field('Rating', rating, FIELD_IDS.rating), field('Notes', textarea('notes'), FIELD_IDS.notes));
+            // D153 (#447): no Rating field; the concept's `#recipe-rating` select is not drawn.
+            form.append(columns, categories, tags, field('Notes', textarea('notes'), FIELD_IDS.notes));
         } else form.append(el('p', NEGATIVE_HELP, 'small'));
         const line = statusLine(), recovery = new Map();
         const actions = el('div', undefined, 'dialog-actions');
@@ -230,9 +228,7 @@ export function createLibraryEditor({api, parentScope, onCommitted = () => {}, o
         form.append(line.note, actions);
 
         for (const [name, node] of Object.entries(controls)) {
-            node.addEventListener(node.tagName === 'SELECT' ? 'change' : 'input', () => {
-                controller.setField(name, name === 'rating' && node.value === '0' ? null : node.value);
-            }, {signal: views.signal});
+            node.addEventListener('input', () => controller.setField(name, node.value), {signal: views.signal});
         }
         form.addEventListener('submit', event => {
             event.preventDefault();
@@ -246,7 +242,7 @@ export function createLibraryEditor({api, parentScope, onCommitted = () => {}, o
         let categoryKey = null, refocus = null;
         function render(state) {
             for (const [name, node] of Object.entries(controls)) {
-                const value = name === 'rating' ? String(state.draft.rating ?? 0) : state.draft[name];
+                const value = state.draft[name];
                 if (node.value !== value) node.value = value;
             }
             const key = state.categories.join('\n');
@@ -274,7 +270,8 @@ export function createLibraryEditor({api, parentScope, onCommitted = () => {}, o
 
         body.replaceChildren(form);
         return {controller, views, render, start() {
-            if (draft) for (const [name, value] of Object.entries(draft)) controller.setField(name, value);
+            // A kept draft carries only the editor's own fields (D153: never a rating).
+            if (draft) for (const [name, value] of Object.entries(draft)) if (EDITOR_FIELDS.includes(name)) controller.setField(name, value);
             render(controller.state);
             controller.initialize();
         }};

web/js/workspace/library-layout.mjs

@@ -60,9 +60,3 @@ export function usedLabel(lastUsedAt, now = new Date()) {
     const year = used.getFullYear() === now.getFullYear() ? '' : `, ${used.getFullYear()}`;
     return `Used ${MONTHS[used.getMonth()]} ${used.getDate()}${year}`;
 }
-
-/** The concept's stars (`★` × rating) with their visually hidden reading; an unrated parent shows neither. */
-export function ratingLabel(rating) {
-    if (!Number.isInteger(rating) || rating < 1 || rating > 5) return {stars: '', hidden: ''};
-    return {stars: '★'.repeat(rating), hidden: `Rated ${rating} out of 5`};
-}

web/js/workspace/library-selection.mjs

@@ -0,0 +1,280 @@
+import { createButton } from '../shared/toolkit/core.js';
+import { selectionLabel, toggleSelection } from './images-state.mjs';
+import { applyKeys, rangeKeys } from './images-multiselect.mjs';
+import { CATEGORY_DIALOG, CATEGORY_FAILED, CATEGORY_MAX_LENGTH, setOutcome } from './images-category.mjs';
+import { NO_TAGS, parseBulkTags } from './images-tagging-bulk.mjs';
+import { openInputDialog } from './workspace-dialogs.mjs';
+import { countOf } from './plural.mjs';
+import { showWorkspaceToast } from './toast.mjs';
+
+/**
+ * The Prompt library's selection and its bulk actions (D148, D151, beyond the concept): "Add tags" (#442, D151) and "Set
+ * category" (#437), the legacy admin page's bulk Add Tags and Set Category in the workspace. Built from the concept's
+ * Images parts only: each card's `.image-overlays` checkbox and the `.bulkbar` row ("N selected", the actions in the
+ * Images bar's order, Clear). The selection is a set of
+ * prompt ids over the shelf's loaded items, so windowed-out cards keep their state; Shift+click on a checkbox applies
+ * the anchor's range (the Images helpers, D135). The dialog, the range helpers, the count and the toast text are the
+ * Images bar's own; one bar core shared with images-selection.mjs is a follow-up (D148).
+ *
+ * Apply posts `{prompt_ids, category|tags, expected_epoch}` to `POST /prompt_manager/bulk/category` or `/bulk/tags`:
+ * the epoch the shelf was read under, so a restored storage answers 409 `storage-changed` and nothing is written.
+ * Add tags writes prompt tags only, never image tags or auto-tags.
+ */
+export const CATEGORY_URL = '/prompt_manager/bulk/category';
+export const PROMPT_CATEGORY_DIALOG = Object.freeze({ ...CATEGORY_DIALOG, title: 'Categorize selected prompts' });
+// The route's cap (`BULK_CATEGORY_MAX_PROMPTS`): more opens nothing.
+export const MAX_CATEGORY_PROMPTS = 1000;
+export const TOO_MANY_PROMPTS = 'Select at most 1,000 prompts.';
+export const STORAGE_CHANGED = 'Prompt storage was restored since this page loaded. No category was changed.';
+export const UNCONFIRMED = 'The category change could not be confirmed. The Library was reloaded.';
+export const NONE_FOUND = 'None of the selected prompts could be found. The Library was reloaded.';
+export const CONTROL_CHARACTERS = 'A category cannot contain control characters.';
+export { CATEGORY_FAILED };
+export const TAGS_URL = '/prompt_manager/bulk/tags';
+export const TAG_OPTIONS_URL = '/prompt_manager/tags';
+export const PROMPT_TAGS_DIALOG = Object.freeze({ title: 'Tag selected prompts', label: 'Tags, separated by commas', action: 'Apply' });
+// What the editor's save accepts for one prompt (`validate_tags`): the route refuses more.
+export const MAX_PROMPT_TAGS = 20;
+export const MAX_TAG_LENGTH = 50;
+export const TAG_PROBLEMS = Object.freeze({ none: NO_TAGS, many: 'Enter at most 20 tags.',
+    long: 'A tag can have at most 50 characters.', control: 'A tag cannot contain control characters.' });
+export const TAGS_FAILED = 'Tags could not be added. Try again.';
+export const TAGS_STORAGE_CHANGED = 'Prompt storage was restored since this page loaded. No tags were added.';
+export const TAGS_UNCONFIRMED = 'Adding the tags could not be confirmed. The Library was reloaded.';
+// Answers that may follow a committed write (the Library editor's unknown outcome): no status, or a gateway's.
+const UNCONFIRMED_STATUSES = new Set([502, 503, 504]);
+
+const CHECKBOX = '.image-overlays input[type="checkbox"]';
+const CARD = 'article[data-prompt-id]';
+
+function el(tag, text, className) {
+    const node = document.createElement(tag);
+    if (text !== undefined) node.textContent = text;
+    if (className) node.className = className;
+    return node;
+}
+
+/**
+ * What a settled write means for the page: `toast` to show, whether the write may have `committed` (the page then
+ * invalidates its cached Library views) and whether the shelf must `reload` (which clears the selection).
+ */
+export function writeOutcome(result, failure, kind = 'category') {
+    const copy = WRITES[kind];
+    if (failure === null && result?.success) {
+        const found = ['updated', 'unchanged', 'over_limit'].reduce((sum, key) => sum + (Number(result[key]) || 0), 0);
+        if (!found) return { toast: NONE_FOUND, committed: false, reload: true };
+        return { toast: copy.success(result), committed: (Number(result.updated) || 0) > 0, reload: true };
+    }
+    if (failure?.code === 'storage-changed') return { toast: copy.storage, committed: false, reload: true };
+    const status = failure?.status;
+    if (!Number.isInteger(status) || UNCONFIRMED_STATUSES.has(status)) return { toast: copy.unconfirmed, committed: true, reload: true };
+    return { toast: copy.failed, committed: false, reload: false };
+}
+
+/** The Add tags toast (#442): what was added, then what already had them, was skipped or could not be found. */
+export function tagsOutcome(result) {
+    const updated = Number(result?.updated) || 0, unchanged = Number(result?.unchanged) || 0;
+    const skipped = Number(result?.over_limit) || 0, missing = Number(result?.not_found) || 0;
+    const names = Array.isArray(result?.tags) ? result.tags.join(', ') : '';
+    const parts = [];
+    if (updated) parts.push(`Added “${names}” to ${countOf(updated, 'prompt')}.`);
+    if (unchanged) parts.push(`${countOf(unchanged, 'prompt')} already had ${updated ? 'them' : `“${names}”`}.`);
+    if (skipped) parts.push(`${countOf(skipped, 'prompt')} would pass ${MAX_PROMPT_TAGS} tags and ${skipped === 1 ? 'was' : 'were'} skipped.`);
+    if (missing) parts.push(`${missing.toLocaleString('en-US')} could not be found.`);
+    return parts.join(' ');
+}
+
+const WRITES = Object.freeze({
+    category: { success: result => setOutcome(result, result.category, 'prompt'), storage: STORAGE_CHANGED,
+        unconfirmed: UNCONFIRMED, failed: CATEGORY_FAILED },
+    tags: { success: tagsOutcome, storage: TAGS_STORAGE_CHANGED, unconfirmed: TAGS_UNCONFIRMED, failed: TAGS_FAILED },
+});
+
+/**
+ * "Select all N loaded" (#444 ruling, D151; the legacy page's select-all checkbox): offered while some loaded prompts
+ * are unchecked and no more are loaded than an action accepts. It sits where D135's "Select all N matching" sits on
+ * Images, but that offer appears once every loaded image is checked; this one reaches only the loaded prompts.
+ */
+export function offersLoaded(selected, keys) {
+    return selected.size > 0 && keys.length <= MAX_CATEGORY_PROMPTS && keys.some(key => !selected.has(key));
+}
+export const loadedOffer = count => `Select all ${count.toLocaleString('en-US')} loaded`;
+
+/** The Add tags field's check before Apply: the parsed names, as `validate_tags` would judge them. */
+export function tagsProblem(text) {
+    const tags = parseBulkTags(text);
+    if (!tags.length) return TAG_PROBLEMS.none;
+    if (tags.length > MAX_PROMPT_TAGS) return TAG_PROBLEMS.many;
+    if (tags.some(tag => tag.length > MAX_TAG_LENGTH)) return TAG_PROBLEMS.long;
+    if (tags.some(tag => /[\x00-\x1f]/.test(tag))) return TAG_PROBLEMS.control;
+    return null;
+}
+
+/** The field's check before Apply: what `validate_category` refuses beyond the length the field already limits. */
+export function categoryProblem(value) {
+    return /[\x00-\x1f]/.test(value) ? CONTROL_CHARACTERS : null;
+}
+
+/**
+ * Mount the selection on the shelf `grid`. `getItems()` is the shelf's loaded items in order, `getEpoch()` the
+ * storage epoch they were read under, `getCategories()` the Library's category list (the field's suggestions).
+ * `onCommitted(kind, receipt)` (`category` or `tags`) runs after a write that may have changed prompts, even once the page is gone (the page
+ * invalidates its cached Library views); `onReload()` then reloads the mounted shelf (which calls `clear`). Focus moves
+ * to `focusFallback` when it was on the bar the reload hides.
+ */
+export function mountLibrarySelection({ scope, api, grid, getItems, getEpoch, getCategories, focusFallback, onCommitted,
+    onReload }) {
+    let checked = Object.freeze(new Set()), anchor = null, shiftBox = null, writing = false;
+    const element = el('div');
+    element.id = 'library-bulkbar';
+    const bulk = el('div', undefined, 'bulkbar'), count = el('strong');
+    const addTags = createButton('Add tags');
+    const setCategory = createButton('Set category');
+    const clearButton = createButton('Clear');
+    const selectLoaded = createButton('');
+    const actions = [addTags, setCategory, selectLoaded];
+    bulk.append(count, addTags, setCategory, clearButton);
+    scope.listen(selectLoaded, 'click', () => {
+        if (writing) return;
+        selectAll();
+        // The pressed offer left the bar: focus moves to its first action, as after D135's offer.
+        addTags.focus({ preventScroll: true });
+    });
+    scope.listen(addTags, 'click', () => openTags());
+    scope.listen(setCategory, 'click', () => openCategory());
+    scope.listen(clearButton, 'click', () => { clear(); focusFallback.focus({ preventScroll: true }); });
+    // Escape in the bar clears the selection; the bar hides, so focus moves to the page heading as after Clear.
+    scope.listen(element, 'keydown', event => {
+        if (event.key !== 'Escape' || !checked.size) return;
+        event.preventDefault();
+        clear();
+        focusFallback.focus({ preventScroll: true });
+    });
+
+    const ids = () => getItems().map(item => item.id);
+    function render() {
+        if (!checked.size) { bulk.remove(); return; }
+        const label = selectionLabel(checked.size);
+        if (count.textContent !== label) count.textContent = label;
+        const keys = ids();
+        if (offersLoaded(checked, keys)) {
+            const offer = loadedOffer(keys.length);
+            if (selectLoaded.textContent !== offer) selectLoaded.textContent = offer;
+            if (count.nextElementSibling !== selectLoaded) count.after(selectLoaded);
+        } else selectLoaded.remove();
+        if (bulk.parentNode !== element) element.append(bulk);
+    }
+    function paint() {
+        for (const box of grid.querySelectorAll(CHECKBOX)) {
+            const on = checked.has(Number(box.closest(CARD).dataset.promptId));
+            if (box.checked !== on) box.checked = on;
+        }
+    }
+    function update(next) { checked = next; paint(); render(); }
+    function clear() {
+        anchor = null;
+        if (checked.size) update(Object.freeze(new Set()));
+    }
+    /** Every loaded prompt, windowed-out cards included; the anchor stays. */
+    function selectAll() {
+        update(applyKeys(checked, ids(), true));
+    }
+    // The checkbox toggles natively; a Shift+click applies the anchor's range with the box's new state.
+    scope.listen(grid, 'click', event => {
+        const box = event.target.closest('.image-overlays > label');
+        if (box && event.shiftKey) shiftBox = Number(box.closest(CARD).dataset.promptId);
+    });
+    scope.listen(grid, 'change', event => {
+        if (!event.target.matches(CHECKBOX)) return;
+        const id = Number(event.target.closest(CARD).dataset.promptId), on = event.target.checked;
+        if (shiftBox === id) {
+            const keys = rangeKeys(ids(), anchor, id);
+            if (anchor === null || keys.length === 1) anchor = id;
+            update(applyKeys(checked, keys, on));
+        } else {
+            anchor = id;
+            update(toggleSelection(checked, id, on));
+        }
+        shiftBox = null;
+    });
+
+    /** The card's overlay: the concept's `.image-overlays > label > input`, checked from the selection. */
+    function overlayFor(item, name) {
+        const overlays = el('div', undefined, 'image-overlays'), label = el('label'), box = el('input');
+        box.type = 'checkbox';
+        box.checked = checked.has(item.id);
+        box.setAttribute('aria-label', `Select ${name}`);
+        label.append(box);
+        overlays.append(label);
+        return overlays;
+    }
+
+    /** The selected loaded prompts and the shelf epoch, or null after a toast says why nothing can be sent. */
+    function target(failed) {
+        if (writing) return null;
+        const loaded = new Set(ids()), selected = [...checked].filter(id => loaded.has(id));
+        if (selected.length > MAX_CATEGORY_PROMPTS) { showWorkspaceToast(TOO_MANY_PROMPTS); return null; }
+        const epoch = getEpoch();
+        if (!selected.length || typeof epoch !== 'string') { showWorkspaceToast(failed); return null; }
+        return { prompt_ids: selected, expected_epoch: epoch };
+    }
+    function openCategory() {
+        const chosen = target(CATEGORY_FAILED);
+        if (!chosen) return;
+        openInputDialog({ scope, trigger: setCategory, title: PROMPT_CATEGORY_DIALOG.title, label: PROMPT_CATEGORY_DIALOG.label,
+            action: PROMPT_CATEGORY_DIALOG.action, maxLength: CATEGORY_MAX_LENGTH, suggestions: getCategories(),
+            validate: categoryProblem, onApply: category => write(CATEGORY_URL, { ...chosen, category }, 'category') });
+    }
+    /** Add tags (#442): the tag vocabulary as suggestions, the editor's list; a failed read opens without them. */
+    async function openTags() {
+        const chosen = target(TAGS_FAILED);
+        if (!chosen) return;
+        let suggestions = null;
+        try {
+            const result = await api.get(TAG_OPTIONS_URL, { signal: scope.signal, cache: 'no-store' });
+            if (Array.isArray(result?.tags)) suggestions = result.tags.filter(tag => typeof tag === 'string');
+        } catch {
+            // Suggestions are auxiliary; the field works without them.
+        }
+        if (scope.signal.aborted) return;
+        openInputDialog({ scope, trigger: addTags, ...PROMPT_TAGS_DIALOG, suggestions, validate: tagsProblem,
+            onApply: text => write(TAGS_URL, { ...chosen, tags: parseBulkTags(text) }, 'tags') });
+    }
+    /** The write carries no page signal: once sent it completes even if the page is left. A failure keeps the selection. */
+    async function write(url, body, kind) {
+        writing = true;
+        for (const button of actions) button.setAttribute('aria-disabled', 'true');
+        let result = null, failure = null;
+        try {
+            result = await api.post(url, body);
+        } catch (error) {
+            failure = error;
+        } finally {
+            writing = false;
+            for (const button of actions) button.removeAttribute('aria-disabled');
+        }
+        const outcome = writeOutcome(result, failure, kind);
+        // A left page still invalidates what it cached, so a return never shows the old categories.
+        if (outcome.committed) onCommitted(kind, result?.success ? result : { phase: 'unconfirmed' });
+        if (scope.signal.aborted) return;
+        const focused = element.contains(document.activeElement);
+        showWorkspaceToast(outcome.toast);
+        if (!outcome.reload) return;
+        onReload();
+        if (focused) focusFallback.focus({ preventScroll: true });
+    }
+
+    render();
+    return {
+        element,
+        overlayFor,
+        isChecked: id => checked.has(id),
+        hasSelection: () => checked.size > 0,
+        clear,
+        selectAll,
+        /** A loaded batch: the count and the offer follow it. */
+        listingChanged: () => { if (checked.size) render(); },
+        /** Whether a Set category write is still in flight. */
+        writing: () => writing,
+    };
+}

web/js/workspace/library-state.mjs

@@ -11,7 +11,7 @@ function parameters(filters,limit) {
     if(filters.tags!==undefined && (!Array.isArray(filters.tags) || filters.tags.some(tag=>typeof tag!=='string'))) throw new TypeError('Invalid library tags');
     for(const tag of filters.tags || []) if(tag.trim()) query.append('tag',tag.trim());
     const sort=filters.sort || 'newest';
-    if(!['newest','oldest','name','rating'].includes(sort)) throw new TypeError('Invalid library sort');
+    if(!['newest','oldest','name'].includes(sort)) throw new TypeError('Invalid library sort');
     query.set('sort',sort);return query;
 }
 // `variant_total` is the filtered parents' active negative variants (the concept caption's "M negative variants").
@@ -19,7 +19,7 @@ function validatePage(page,cursor) {
     if(!page || !Array.isArray(page.items) || page.items.some(item=>!parent(item)) || !Number.isSafeInteger(page.total) || page.total<0 || !Number.isSafeInteger(page.variant_total) || page.variant_total<0 || typeof page.has_more!=='boolean' || !(page.next_cursor===null || (typeof page.next_cursor==='string' && page.next_cursor.length)) || page.has_more!==(page.next_cursor!==null) || (page.has_more && cursor===page.next_cursor)) throw new TypeError('Invalid library page');
 }
 export function createLibraryStream({api,onChange=()=>{},restore}) {
-    let state=Object.freeze({items:Object.freeze([]),total:0,variantTotal:null,cursor:null,hasMore:true,loading:false,error:null});
+    let state=Object.freeze({items:Object.freeze([]),total:0,variantTotal:null,cursor:null,hasMore:true,loading:false,error:null,epoch:null});
     let query=parameters({},40),request,disposed=false,seen=new Set();
     if(restore) {
         query=parameters(restore.filters,40);const saved=restore.state;
@@ -27,7 +27,7 @@ export function createLibraryStream({api,onChange=()=>{},restore}) {
         const firstPagePending=saved?.variantTotal===null && Array.isArray(saved.items) && !saved.items.length && saved.total===0 && saved.cursor===null && saved.hasMore===true;
         if(!firstPagePending) validatePage({items:saved.items,total:saved.total,variant_total:saved.variantTotal,next_cursor:saved.cursor,has_more:saved.cursor!==null},null);
         if(new Set(saved.items.map(item=>item.id)).size!==saved.items.length) throw new TypeError('Duplicate restored parent');
-        state=Object.freeze({...saved,items:Object.freeze(saved.items.map(item=>Object.freeze({...item}))),loading:false,error:null});seen=new Set(state.items.map(item=>item.id));
+        state=Object.freeze({...saved,epoch:saved.epoch ?? null,items:Object.freeze(saved.items.map(item=>Object.freeze({...item}))),loading:false,error:null});seen=new Set(state.items.map(item=>item.id));
     }
     const emit=patch=>{state=Object.freeze({...state,...patch});onChange(state);};
     async function loadMore() {
@@ -37,14 +37,19 @@ export function createLibraryStream({api,onChange=()=>{},restore}) {
         try {
             const page=await api.get(`/prompt_manager/library?${params}`,{signal:current.signal});
             if(disposed || current!==request) return 0;
-            validatePage(page,cursor);const incoming=[];const added=new Set();
+            validatePage(page,cursor);
+            // #437: the storage epoch the shelf's items were read under, which a bulk write sends as its fence; a later
+            // page from a restored storage is refused (its items would belong to another database).
+            const epoch=typeof page.library_epoch==='string'?page.library_epoch:null;
+            if(state.epoch!==null && epoch!==state.epoch) throw new TypeError('Prompt storage changed while the library was loading');
+            const incoming=[];const added=new Set();
             for(const item of page.items) if(!seen.has(item.id) && !added.has(item.id)) {added.add(item.id);incoming.push(Object.freeze({...item}));}
             for(const value of added) seen.add(value);
-            request=null;emit({items:incoming.length?Object.freeze([...state.items,...incoming]):state.items,total:page.total,variantTotal:page.variant_total,cursor:page.next_cursor,hasMore:page.has_more,loading:false,error:null});return incoming.length;
+            request=null;emit({items:incoming.length?Object.freeze([...state.items,...incoming]):state.items,epoch:state.epoch ?? epoch,total:page.total,variantTotal:page.variant_total,cursor:page.next_cursor,hasMore:page.has_more,loading:false,error:null});return incoming.length;
         } catch(error) {if(!disposed && current===request) {request=null;emit({loading:false,error});}return 0;}
     }
     return {get state(){return state;},loadMore,async reset(filters={}) {
-        if(disposed) return 0;const next=parameters(filters,40);request?.abort();request=null;query=next;seen=new Set();emit({items:Object.freeze([]),total:0,variantTotal:null,cursor:null,hasMore:true,loading:false,error:null});return loadMore();
+        if(disposed) return 0;const next=parameters(filters,40);request?.abort();request=null;query=next;seen=new Set();emit({items:Object.freeze([]),total:0,variantTotal:null,cursor:null,hasMore:true,loading:false,error:null,epoch:null});return loadMore();
     },dispose(){disposed=true;request?.abort();request=null;}};
 }
 /**

web/js/workspace/library.mjs

@@ -6,7 +6,7 @@ import {createLibraryStream,createLibraryDetail,libraryPromptRoles} from './libr
 import {promptRoles} from './prompt-blocks.mjs';
 import {createLibraryEditor,showWorkspaceToast} from './library-editor.mjs';
 import {createLibraryEditorController} from './library-editor-state.mjs';
-import {layoutLibraryRows,visibleLibraryRows,usedLabel,ratingLabel} from './library-layout.mjs';
+import {layoutLibraryRows,visibleLibraryRows,usedLabel} from './library-layout.mjs';
 import {createWorkspaceIcon} from './icons.mjs';
 import {mountRecordedHistory} from './history.mjs';
 import {countOf} from './plural.mjs';
@@ -15,6 +15,7 @@ import {pageHeading} from './heading.mjs';
 import {mountToolbar} from './toolbar.mjs';
 import {libraryFilters,libraryViewState} from './saved-views.mjs';
 import {openRefineDialog,openContentDialog} from './workspace-dialogs.mjs';
+import {mountLibrarySelection} from './library-selection.mjs';
 import {fallBackToOriginal,replaceOnError,showsThumbnail,thumbnailSource} from './thumbnail-support.mjs';
 
 let shelfVisit=null;
@@ -28,8 +29,10 @@ const selectedPairs=new Map();
 // Parent metadata a reuse did not apply, kept for an explicit Edit metadata action.
 const metadataDrafts=new Map();
 const archivedVisible=new Map();
-// The concept's three sorts first, then the Library's appended Oldest first (D14).
-const LIBRARY_SORTS=Object.freeze([['newest','Recently added'],['name','Name A–Z'],['rating','Highest rated'],['oldest','Oldest first']]);
+// The concept's sorts first, then the Library's appended Oldest first (D14); D153 (#447) drops "Highest rated".
+const LIBRARY_SORTS=Object.freeze([['newest','Recently added'],['name','Name A–Z'],['oldest','Oldest first']]);
+// A card's focusable controls in order: the D148 selection checkbox, then its buttons (#437).
+const CARD_CONTROLS='button, input[type="checkbox"]';
 const CATEGORIES_FAILED='Categories could not be loaded. Open the Prompt library again to retry.';
 const ARCHIVED_MESSAGE='Archived. Find it under Show archived in Negative variants.';
 function el(tag,text,className) {
@@ -108,7 +111,7 @@ function outcomeMessage(result,outcome) {
     if(kind!=='create' && kind!=='add-negative') return null;
     if(receipt.positive_created) return 'Saved.';
     if(receipt.pair_created) return kind==='add-negative'?'Negative saved without adding another positive or gallery.'
-        :outcome?.metadataNotApplied?'Saved as a new negative variant of an existing prompt. The title, category, tags, rating and notes you entered were not applied.'
+        :outcome?.metadataNotApplied?'Saved as a new negative variant of an existing prompt. The title, category, tags and notes you entered were not applied.'
             :'Saved as a new negative variant of an existing prompt.';
     if(receipt.pair.archived) return 'This combination is an archived negative variant. It stays archived; restore it from Show archived in Negative variants.';
     return 'This prompt pair already exists. Showing the saved version.';
@@ -119,7 +122,7 @@ function takeSavedSelection(positiveId) {
 }
 function rememberUnappliedMetadata(result,{outcome,draft}={}) {
     if(result.commit.kind!=='create' || !outcome?.metadataNotApplied || !draft) return;
-    const {title:name,category,tags,rating,notes}=draft;metadataDrafts.set(result.detail.positive.id,{title:name,category,tags,rating,notes});
+    const {title:name,category,tags,notes}=draft;metadataDrafts.set(result.detail.positive.id,{title:name,category,tags,notes});
 }
 /** New prompt, from Library or Images: a saved prompt always opens its own history, as in the concept. */
 function createNewPromptEditor(api,scope,main) {
@@ -156,6 +159,8 @@ export function mountLibrary(main,api,parentScope,{positiveId,savedState=null,on
         // An invalidated visit keeps only filters and anchor: the stream restarts and heights are measured again.
         const restorable=visit?.state?visit:null;
         const filters={text:'',category:'',tags:[],sort:'newest',...visit?.filters,...saved};
+        // D153: a sort the shelf no longer offers (an older "Highest rated" visit) reads as Recently added.
+        if(!LIBRARY_SORTS.some(([value])=>value===filters.sort)) filters.sort='newest';
         const section=el('section',undefined,'library-shelf');section.setAttribute('aria-label','Prompt library');
         // The shared concept toolbar (app.js toolbar(), a347–360): Tags live in "Refine this view" (D14), no Refresh (D15).
         // Its own scope, disposed after the visit is saved: removing it first would shorten the page and clamp scrollY.
@@ -171,6 +176,11 @@ export function mountLibrary(main,api,parentScope,{positiveId,savedState=null,on
         // The count is announced when it changes; a reload keeps the last count shown, so nothing else re-announces it.
         captionCount.setAttribute('role','status');
         const grid=el('div',undefined,'library-grid');grid.setAttribute('role','list');grid.setAttribute('aria-label','Positive prompts');
+        // #437 D148: select prompts and set their category (the legacy admin's Set Category); a write reloads the shelf.
+        let categoryNames=[];
+        const selection=mountLibrarySelection({scope,api,grid,getItems:()=>state?.items ?? [],getEpoch:()=>state?.epoch ?? null,
+            getCategories:()=>categoryNames,focusFallback:{focus:()=>focusHeading(main)},
+            onCommitted:(kind,receipt)=>invalidateLibraryAfterWrite({kind:`bulk-${kind}`,receipt}),onReload:()=>{loadCategories();reset();}});
         const message=el('p','','library-status');message.setAttribute('role','status');
         const continuation=el('div',undefined,'library-continuation');
         const more=createButton('Load more prompts');const retry=createButton('Retry prompt library');retry.hidden=true;
@@ -201,7 +211,7 @@ export function mountLibrary(main,api,parentScope,{positiveId,savedState=null,on
         scope.listen(exportButton,'click',exportLibrary);
         headingDescription(main,section,'One positive. Reusable negatives. Every experiment stays connected.');
         // Nothing follows the grid in the concept; the legacy Library stays one click away on Overview (spec §5).
-        continuation.append(message,retry,more);section.append(optionsStatus,exportStatus,caption,grid,continuation);main.append(section);
+        continuation.append(message,retry,more);section.append(optionsStatus,exportStatus,selection.element,caption,grid,continuation);main.append(section);
         const cards=new Map(), rowNodes=new Map(), spacers=new Map(), heights=new Map(), pendingMeasurements=new Map();
         let state,frame=0,loadFrame=0,observer,optionsRequest,focusOverride=-1;
         const columns=()=>window.innerWidth>=1700?3:window.innerWidth<=700?1:2;
@@ -212,7 +222,7 @@ export function mountLibrary(main,api,parentScope,{positiveId,savedState=null,on
         if(restorable?.layout?.inlineSize===inlineSize && restorable.layout.columns===columnCount) for(const [id,height] of restorable.layout.heights) heights.set(id,height);
         const hiddenSave=restorable?null:savedParentId;savedParentId=null;
         function removeCard(card) {measurementObserver?.unobserve(card.node);pendingMeasurements.delete(Number(card.node.dataset.promptId));card.blocks.dispose();card.node.remove();}
-        /** The concept's prompt card (app.js renderLibrary a586) with real counts, last use and rating (#249 D23). */
+        /** The concept's prompt card (app.js renderLibrary a586) with real counts and last use (#249 D23; no rating, D153). */
         function cardFor(item,index) {
             const article=el('article',undefined,'prompt-card');article.dataset.promptId=item.id;article.setAttribute('role','listitem');
             const art=el('div',undefined,'card-art');art.setAttribute('aria-hidden','true');
@@ -238,15 +248,13 @@ export function mountLibrary(main,api,parentScope,{positiveId,savedState=null,on
             const vocabulary=el('div',undefined,'tags');for(const tag of item.tags || []) vocabulary.append(el('span',tag,'chip'));
             content.append(heading,prompts,meta,vocabulary);
             const footer=el('div',undefined,'card-footer');
-            const {stars,hidden}=ratingLabel(item.rating), used=usedLabel(item.last_used_at);
-            const facts=el('span');
-            if(stars) {const shown=el('span',stars);shown.setAttribute('aria-hidden','true');facts.append(shown,el('span',hidden,'ca-visually-hidden'),` \u00b7 ${used}`);}
-            else facts.textContent=used;
+            // D153 (#447): the footer reads only the last use; the concept's rating stars are not drawn.
+            const facts=el('span',usedLabel(item.last_used_at));
             const actions=el('div',undefined,'row-actions');
             const send=createButton('','send icon-btn');send.setAttribute('aria-label',`Send positive prompt to ComfyUI: ${name}`);send.append(createWorkspaceIcon('send'));send.dataset.sendPositive='';
             const explore=createButton('');explore.append(createWorkspaceIcon('branch'),' Explore history');explore.dataset.history='';
             actions.append(send,explore);footer.append(facts,actions);
-            article.append(art,content,footer);article.dataset.index=index;return {node:article,blocks};
+            article.append(selection.overlayFor(item,name),art,content,footer);article.dataset.index=index;return {node:article,blocks};
         }
         function captureAnchor() {
             // Mounted rows can already differ from estimates before their observer batch.
@@ -340,10 +348,11 @@ export function mountLibrary(main,api,parentScope,{positiveId,savedState=null,on
             if(changed) {pendingAnchor ||= anchor;schedule();}
         }):null;
         measurementObserver?.observe(grid);
-        function focusCard(index,last=false) {
+        /** Focus a card's first or `last` control (Tab across the window edge), or its title (`title`: Home/End/arrows). */
+        function focusCard(index,last=false,title=false) {
             if(index<0 || index>=state.items.length) return;
             if(dirty) {cancelAnimationFrame(frame);flush();}focusOverride=index;renderWindow();const card=cards.get(state.items[index].id).node;
-            const controls=card.querySelectorAll('button');(last?controls[controls.length-1]:controls[0]).focus();card.scrollIntoView({block:'nearest'});focusOverride=-1;renderWindow();
+            const controls=card.querySelectorAll(CARD_CONTROLS);(title?card.querySelector('.card-title > button'):last?controls[controls.length-1]:controls[0]).focus();card.scrollIntoView({block:'nearest'});focusOverride=-1;renderWindow();
         }
         // The title and Explore history open the prompt's history; the footer send sends the full positive (D23).
         scope.listen(grid,'click',event=>{
@@ -355,15 +364,19 @@ export function mountLibrary(main,api,parentScope,{positiveId,savedState=null,on
         scope.listen(grid,'keydown',event=>{
             const card=event.target.closest('article[data-prompt-id]');if(!card) return;
             const index=Number(card.dataset.index);
-            if(event.target.matches('[data-history]') && ['Home','End','ArrowLeft','ArrowRight'].includes(event.key)) {event.preventDefault();focusCard(event.key==='Home'?0:event.key==='End'?state.items.length-1:index+(event.key==='ArrowLeft'?-1:1));}
+            // D151 (#444 ruling): Ctrl/Cmd+A selects every loaded prompt and Escape clears, as on Images (focus stays).
+            if((event.ctrlKey || event.metaKey) && !event.altKey && event.key.toLowerCase()==='a') {event.preventDefault();selection.selectAll();return;}
+            if(event.key==='Escape' && selection.hasSelection()) {event.preventDefault();selection.clear();return;}
+            if(event.target.matches('[data-history]') && ['Home','End','ArrowLeft','ArrowRight'].includes(event.key)) {event.preventDefault();focusCard(event.key==='Home'?0:event.key==='End'?state.items.length-1:index+(event.key==='ArrowLeft'?-1:1),false,true);}
             else if(event.key==='Tab') {
-                const controls=card.querySelectorAll('button');const next=index+(event.shiftKey?-1:1);
+                const controls=card.querySelectorAll(CARD_CONTROLS);const next=index+(event.shiftKey?-1:1);
                 const edge=event.shiftKey?controls[0]:controls[controls.length-1];
                 if(event.target===edge && state.items[next] && !cards.has(state.items[next].id)) {event.preventDefault();focusCard(next,event.shiftKey);}
             }
         });
         function render(next) {
-            if(items!==next.items) {
+            const batch=items!==next.items;
+            if(batch) {
                 pendingAnchor ||= captureAnchor();items=next.items;indexesById=new Map(items.map((item,index)=>[item.id,index]));dirty=true;
                 if(!items.length) {
                     anchor=null;pendingAnchor=null;heights.clear();layout={rows:[],height:0};
@@ -372,6 +385,8 @@ export function mountLibrary(main,api,parentScope,{positiveId,savedState=null,on
                 } else if(retainedAnchor) {pendingAnchor=retainedAnchor;retainedAnchor=null;}
             }
             state=next;
+            // A loaded batch: the selection bar's count and Select all N loaded follow it (D151).
+            if(batch) selection.listingChanged();
             if(next.variantTotal!==null) {
                 const count=`${countOf(next.total,'positive prompt')} · ${countOf(next.variantTotal,'negative variant')}`;
                 if(captionCount.textContent!==count) captionCount.textContent=count;caption.hidden=false;
@@ -392,10 +407,10 @@ export function mountLibrary(main,api,parentScope,{positiveId,savedState=null,on
             optionsRequest?.abort();const request=optionsRequest=new AbortController();
             try {
                 const data=await api.get('/prompt_manager/library/options',{signal:request.signal});if(scope.signal.aborted || optionsRequest!==request) return;
-                toolbar.setCategories([...data.categories,filters.category]);toolbar.setValue({category:filters.category});optionsStatus.hidden=true;
+                categoryNames=[...data.categories];toolbar.setCategories([...data.categories,filters.category]);toolbar.setValue({category:filters.category});optionsStatus.hidden=true;
             } catch {if(!scope.signal.aborted && optionsRequest===request) {optionsStatus.textContent=CATEGORIES_FAILED;optionsStatus.hidden=false;}}
         }
-        const reset=()=>{restoreScrollY=null;toolbar.setActiveFilters(refinements());return stream.reset(filters);};
+        const reset=()=>{restoreScrollY=null;selection.clear();toolbar.setActiveFilters(refinements());return stream.reset(filters);};
         /** A toolbar report; a debounced search that lands after a select already applied the same values reloads nothing. */
         function applyToolbar(value) {
             const next={text:value.q.trim(),category:value.category,sort:value.sort};

web/js/workspace/local-resolver.mjs

@@ -4,7 +4,7 @@
 // tests/test_local_inspect_parity.py compares every answer with the server's on the same records. Values follow
 // Python's model (py-json.mjs): dict order, int versus float, code-point lengths, `str.strip()`.
 //
-// Task 8 ported execution evidence and graph tracing for one sampler; Task 9 adds candidate selection among several
+// Task 8 ported execution evidence and graph tracing for one sampler; Task 9 added candidate selection among several
 // samplers and detailer stages (local-resolver-stages.mjs) and A1111 `parameters` (local-parameters.mjs). No DOM.
 import { PyValueError, isDict, isList, isInt, pyJsonLoads, pyLen, pySlice, pyStrip, pyStr, pyGet } from './py-json.mjs';
 import {

web/js/workspace/media-viewer.mjs

@@ -19,7 +19,12 @@ function element(tag, className, text) {
     return node;
 }
 
-export function createMediaViewer({api,container,parentScope,inspectionController,sender,onSelect,onRefresh,onRestoreFocus,onClose,onMediaError,adapter = rootedFileAdapter}) {
+/**
+ * #331 D139: with the client store `collections`, the viewer's inspector offers "Collections…"; the changes it confirms
+ * reach `onCollectionsChanged({added, removed})` once the viewer closes, so the viewer keeps its place while open.
+ */
+export function createMediaViewer({api,container,parentScope,inspectionController,sender,onSelect,onRefresh,onRestoreFocus,onClose,onMediaError,adapter = rootedFileAdapter,
+    collections = null,onCollectionsChanged = null}) {
     const scope = createScope();
     let active = null, retryVersion = 0;
     const dispose = () => scope.dispose();
@@ -36,7 +41,7 @@ export function createMediaViewer({api,container,parentScope,inspectionControlle
         active?.close(false);
         const {items,selectedKey,trigger = document.activeElement} = options;
         let mounted = false, initial, entry, currentKey, viewer = null, withoutViewer = false, mediaScope = null, closed = false;
-        let originalDownload = null, inspectionState = inspectionController.state;
+        let originalDownload = null, inspectionState = inspectionController.state, collectionChanges = null;
         const session = createViewerSession({adapter,items,initialKey:selectedKey,onChange:(state,selected) => {
             if (!mounted) { initial = [state,selected]; return; }
             update(state,selected);
@@ -99,7 +104,11 @@ export function createMediaViewer({api,container,parentScope,inspectionControlle
         const localSender = {open(roles,trigger) {session.stop();sender.open(roles,trigger);}};
         const inspector = mountFileInspector(details,{api,sender:localSender,parentScope:local,variant:'full',onInteraction:() => session.stop(),onRetry:() => inspectionController.retry(),
             onLoad:() => inspectionController.load?.(),
-            onRefresh:onRefresh ? () => { closeViewer();onRefresh(); } : undefined});
+            onRefresh:onRefresh ? () => { closeViewer();onRefresh(); } : undefined,collections,
+            onCollectionsChanged:({added,removed}) => {
+                collectionChanges ??= {added:[],removed:[]};
+                collectionChanges.added.push(...added);collectionChanges.removed.push(...removed);
+            }});
         function closeViewer(restore = true) {
             if (closed) return;
             closed = true;active = null;local.dispose();
@@ -108,6 +117,7 @@ export function createMediaViewer({api,container,parentScope,inspectionControlle
                 else onRestoreFocus?.(entry);
             }
             if (!scope.signal.aborted) onClose?.();
+            if (collectionChanges && !scope.signal.aborted) onCollectionsChanged?.(collectionChanges);
         }
         active = {close:closeViewer,stop:() => session.stop()};
         local.cleanup(() => {

web/js/workspace/saved-views.mjs

@@ -6,7 +6,8 @@
  *   view", "View name", `Search: <q>` or "My workspace view", Cancel / Save; an empty name is "Saved view".
  * - The sidebar SAVED VIEWS group lists each view (concept `nav()` a331–335) as `#/saved/<id>` (a1158–1162).
  * - A view's state always holds every key of its page (the API requires them all); a value its page no longer
- *   accepts is dropped to that key's default, so an opened view applies what it can.
+ *   accepts is dropped to that key's default, so an opened view applies what it can (a view saved under the retired
+ *   "Highest rated" sort opens on Recently added, D153).
  * - D51: an open view's heading Edit (maintainer ruling, beyond the concept): "Edit saved view" with "View name",
  *   Delete… (danger), Cancel and Save; Delete… asks inside the same dialog, never through `window.confirm`.
  */
@@ -36,7 +37,7 @@ const tags = { fallback: [], accepts: value => Array.isArray(value) && value.len
     typeof tag === 'string' && tag.trim() && tag.length <= MAX_TAG_LENGTH && !tag.includes(',') && !CONTROL.test(tag)) };
 const FIELDS = Object.freeze({
     images: Object.freeze({
-        q: text(500), category: text(100), sort: choice(['newest', 'oldest', 'name', 'name-desc', 'rating'], 'newest'),
+        q: text(500), category: text(100), sort: choice(['newest', 'oldest', 'name', 'name-desc'], 'newest'),
         source: { fallback: 'all', accepts: value => typeof value === 'string' && /^(all|0|[1-9][0-9]{0,8})$/.test(value) },
         folder: text(1024), media: choice(['all', 'image', 'video'], 'all'), tags,
         // A view saved on Favorites (#249 D51) reopens Images filtered to favorites.
@@ -45,7 +46,7 @@ const FIELDS = Object.freeze({
         collection: { fallback: '', accepts: value => typeof value === 'string' && /^(|[1-9][0-9]{0,8})$/.test(value) },
     }),
     library: Object.freeze({
-        q: text(500), category: text(100), sort: choice(['newest', 'oldest', 'name', 'rating'], 'newest'), tags,
+        q: text(500), category: text(100), sort: choice(['newest', 'oldest', 'name'], 'newest'), tags,
     }),
 });
 

web/js/workspace/system-backups.mjs

@@ -326,6 +326,43 @@ export function onRestoreFinished(activity, onFinished) {
     });
 }
 
+const OPEN_DIALOG = 'dialog[open]';
+const DIALOG_MUTATIONS = Object.freeze({ subtree: true, childList: true, attributes: true, attributeFilter: ['open'] });
+
+/**
+ * #415: the S9 reload after a finished restore never runs under an open dialog, which it would close with whatever
+ * was typed in it. `request()` runs `reload()` at once when no dialog is open; otherwise it runs once the last open
+ * dialog has closed (repeated requests meanwhile run it once). The dialog keeps its entries, and its next write is
+ * refused by the restored storage with its own note and Reload, as when this page never saw the restore. A
+ * MutationObserver sees every dialog close: removing `open` and removing the node are synchronous, while the dialog
+ * `close` event is a later task that never bubbles, and every dialog here is removed in the task that closes it.
+ * `dispose()` drops a waiting reload. Returns `{request, dispose}`.
+ */
+export function createRestoreReload(reload, { doc = globalThis.document, Observer = globalThis.MutationObserver } = {}) {
+    let observer = null;
+    const dialogOpen = () => doc.querySelector(OPEN_DIALOG) !== null;
+    const stop = () => {
+        observer?.disconnect();
+        observer = null;
+    };
+    return {
+        request() {
+            if (observer) return;
+            if (!dialogOpen()) {
+                reload();
+                return;
+            }
+            observer = new Observer(() => {
+                if (dialogOpen()) return;
+                stop();
+                reload();
+            });
+            observer.observe(doc.documentElement, DIALOG_MUTATIONS);
+        },
+        dispose: stop,
+    };
+}
+
 function surface(heading, ...nodes) {
     const section = el('section', undefined, 'surface');
     section.append(heading, ...nodes);

web/js/workspace/system-overview.mjs

@@ -11,7 +11,8 @@
  *   D99: its Review thumbnails opens the same "Thumbnail maintenance" dialog as Sources (thumbnail-review.mjs).
  *   #317: it reads "N missing · M outdated · K failed"; while previews are outdated (another format than the
  *   configured one) a second button, "Convert outdated", opens that dialog on its Convert outdated scope (D131).
- * - D96: the Statistics dialog lists the real runs and average rating per category, then "Library totals".
+ * - D96: the Statistics dialog lists the real runs per category, then "Library totals" (D153: no rating column or
+ *   rated-images total).
  * - D95: the "Health report" lists every check the server ran, Missing files per image source, with a chip per real
  *   result and when it was checked. The check runs from Maintenance's notice ("Run health checks", D123) as a
  *   read job of the registry, tracked in the Activity store (its event stream); the report opens when it ends.
@@ -168,21 +169,16 @@ export function healthSummary(report) {
     return `${time}Nothing was changed; review the checks with issues.`;
 }
 
-/** D96: the concept's "Runs by category" rows `[category, runs, average rating]`. */
+/** D96: the concept's "Runs by category" rows `[category, runs]`; D153 drops its average rating column. */
 export function categoryRows(statistics) {
-    return statistics.by_category.map(row => [
-        row.category ?? 'Uncategorized',
-        number(row.runs),
-        row.average_rating === null || row.average_rating === undefined ? 'Not rated'
-            : `${Number(row.average_rating).toLocaleString(undefined, { maximumFractionDigits: 2 })} / 5`,
-    ]);
+    return statistics.by_category.map(row => [row.category ?? 'Uncategorized', number(row.runs)]);
 }
 
-/** D96: the added "Library totals" rows `[label, count]` (the legacy statistics and the rated images). */
+/** D96: the added "Library totals" rows `[label, count]` (the legacy statistics; D153 drops the rated images). */
 export function totalsRows(statistics) {
     const { totals } = statistics;
-    return [['Prompts', totals.total_prompts], ['Tags', totals.total_tags], ['Images', totals.total_images],
-        ['Rated images', statistics.rated_images]].map(([label, value]) => [label, number(value)]);
+    return [['Prompts', totals.total_prompts], ['Tags', totals.total_tags], ['Images', totals.total_images]]
+        .map(([label, value]) => [label, number(value)]);
 }
 
 function table(caption, head, rows, empty) {
@@ -215,7 +211,7 @@ function table(caption, head, rows, empty) {
 /** The Statistics dialog: "Runs by category" (the concept's table), then "Library totals" (D96). */
 export function statisticsNodes(statistics) {
     return [
-        table('Runs by category', ['Category', 'Runs', 'Avg. rating'], categoryRows(statistics), 'No prompts in the library yet.'),
+        table('Runs by category', ['Category', 'Runs'], categoryRows(statistics), 'No prompts in the library yet.'),
         table('Library totals', ['Total', 'Count'], totalsRows(statistics), ''),
     ];
 }
@@ -378,7 +374,7 @@ function overviewNodes(overview, { api, activity, notify, scope, navigate, openS
         operationRow({ scope, title: 'One display configuration', description: 'Grid and density settings are shared across views.',
             label: 'Appearance', onClick: go('appearance') }),
         operationRow({ scope, title: sourcesTitle, description: sourcesDescription, label: 'Manage sources', onClick: go('sources') }),
-        operationRow({ scope, title: 'Statistics', description: 'Runs and ratings by category, with library totals.', label: 'Explore',
+        operationRow({ scope, title: 'Statistics', description: 'Runs by category, with library totals.', label: 'Explore',
             onClick: openStatistics }),
     ]);
     const columns = el('div', undefined, 'two-col');

web/js/workspace/toolbar.mjs

@@ -13,6 +13,7 @@ export const SEARCH_DELAY = 250;
 
 // The concept's sort labels, in the concept's order. Matching is by exact label (note the en dash in "Name A–Z"); any
 // other label is one of the page's appended options.
+// No page passes "Highest rated" since D153 (#447); it stays here only to keep the concept's order.
 const CONCEPT_SORT_LABELS = Object.freeze(['Recently added', 'Name A–Z', 'Highest rated']);
 
 const isText = value => typeof value === 'string';

web/js/workspace/workspace-dialogs.mjs

@@ -112,13 +112,13 @@ function bindScope(scope) {
     });
 }
 
-function show(heading, nodes, { scope, trigger, focusTarget, onClose = null, canClose = null }) {
+function show(heading, nodes, { scope, trigger, focusTarget, onClose = null, canClose = null, review = false }) {
     shell ??= createShell();
     if (!shell.dialog.isConnected) document.body.append(shell.dialog);
     bindScope(scope);
     // A swapped-out body is gone from the shell, so its owner releases it.
     const replaced = session?.onClose;
-    session = { scope, trigger: trigger ?? session?.trigger ?? null, onClose, canClose };
+    session = { scope, trigger: trigger ?? session?.trigger ?? null, onClose, canClose, review };
     replaced?.();
     shell.title.textContent = heading;
     shell.body.replaceChildren(...nodes);
@@ -133,6 +133,14 @@ export function closeWorkspaceDialog() {
     shell = null;
 }
 
+/**
+ * #415: a finished restore ends an open record-changing review (`openReviewDialog`) as a dismissal, since the plan it
+ * reviewed no longer describes storage and it holds nothing typed. Any other dialog stays open.
+ */
+export function closeReviewDialog() {
+    if (session?.review) closeWorkspaceDialog();
+}
+
 function fillSelect(select, key, options) {
     if (!Array.isArray(options) || options.length === 0 ||
         options.some(pair => !Array.isArray(pair) || typeof pair[0] !== 'string' || typeof pair[1] !== 'string')) {
@@ -324,8 +332,9 @@ export const INPUT_SUGGESTIONS_ID = 'workspace-input-suggestions';
  * reports `fallback` when one is given (Save view's "Saved view", a1339). A positive `maxLength` limits the field.
  * `validate(value)` may refuse a value with a message: the dialog then stays open with the typed text and the field
  * reports the message (its native validity) until it is edited. `suggestions` (text list) offers values in a
- * `<datalist>` under the field; `secondary` (`{label, onClick}`) adds a button between Cancel and the action that
- * closes the dialog and runs `onClick` (#332).
+ * `<datalist>` under the field (Set category's categories, #332; Add to collection's names, #331 D138); the field
+ * still takes any text. `secondary` (`{label, onClick}`) adds a button between Cancel and the action that closes the
+ * dialog and runs `onClick` (#332).
  */
 export function openInputDialog({ scope, title, label, action, value = '', maxLength, fallback, onApply, trigger = null,
     validate = null, suggestions = null, secondary = null }) {
@@ -392,13 +401,13 @@ export function openInputDialog({ scope, title, label, action, value = '', maxLe
  * another dialog replaces its body, so the page can dispose what it mounted there. While `canClose()` answers false,
  * × and Escape leave the dialog open (the page, a route change or a replacing dialog still close it).
  */
-export function openContentDialog({ scope, title, nodes, onClose = null, trigger = null, canClose = null }) {
+export function openContentDialog({ scope, title, nodes, onClose = null, trigger = null, canClose = null, focusTarget = null }) {
     requireScope(scope);
     if (typeof title !== 'string' || !title) throw new TypeError('The dialog needs a title');
     if (!Array.isArray(nodes) || nodes.some(node => !node || typeof node !== 'object')) throw new TypeError('The dialog needs its nodes');
     if (onClose !== null && typeof onClose !== 'function') throw new TypeError('The dialog onClose must be a function');
     if (canClose !== null && typeof canClose !== 'function') throw new TypeError('The dialog canClose must be a function');
-    show(title, nodes, { scope, trigger, onClose, canClose });
+    show(title, nodes, { scope, trigger, onClose, canClose, focusTarget });
 }
 
 export const SCAN_REVIEW_SENTENCE = 'Enrich positive / negative associations, reuse canonical text and report unchanged files. ' +
@@ -497,7 +506,7 @@ export function openReviewDialog({ scope, title, heading = null, grid, options =
     actions.append(cancel, start);
     nodes.push(actions);
     // Cancel, ×, Escape, a replacing dialog or the page leaving: everything but the primary is a dismissal.
-    show(title, nodes, { scope, trigger, onClose: onDismiss ? () => { if (!started) onDismiss(); } : null });
+    show(title, nodes, { scope, trigger, onClose: onDismiss ? () => { if (!started) onDismiss(); } : null, review: true });
     return {
         setValue(term, text) {
             const dd = values.get(term);

web/js/workspace/workspace.css

@@ -169,6 +169,13 @@ main:focus-visible { outline: 2px solid var(--ca-focus); outline-offset: 4px; }
 .masonry-grid .image-open video { aspect-ratio: var(--media-ratio); object-fit: contain; }
 .list-grid .image-open video { height: 95px; aspect-ratio: auto; }
 
+/* #451 D157: a cover crop of a generated image keeps its top (heads, faces) instead of its middle. The vertical 20%
+   acts only when the image is taller than its box; a wider image overflows sideways and stays centred. Masonry shows
+   the whole image (contain) and keeps the centre. */
+.image-grid:not(.masonry-grid) .image-open :is(img, video), .card-art img, .detail-cover, .tag-image-preview {
+    object-position: 50% 20%; /* D157 */
+}
+
 /* Images viewer (#249 C6): the concept's `#viewer-dialog` shell (s1008–1058, ≤700 s1456–1474). #249 D34: real media,
    titles and states that the concept's one demo image never shows. */
 /* #249 D34: a real title can wrap (long linked titles and filenames). Above 700px the dialog stacks its header and the
@@ -230,6 +237,11 @@ main:focus-visible { outline: 2px solid var(--ca-focus); outline-offset: 4px; }
    details, and an unavailable download stays visibly inert next to its reason. */
 .file-inspector { min-inline-size: 0; overflow-wrap: anywhere; }
 .file-inspector [aria-disabled="true"] { opacity: .6; cursor: not-allowed; }
+/* #449 D154: the image's own tags and the linked prompt's or recipe's tags sit in the prompt sections' frame (concept
+   .prompt-section), under a muted .role-label; the heading's 8px spaces the chips, as it spaces a prompt's text. */
+.file-inspector .tag-section { border-top: 1px solid var(--line); padding-top: 15px; margin-top: 15px; }
+.file-inspector .tag-section .role-label { color: var(--ca-muted); }
+.file-inspector .tag-section > .tags { margin-top: 0; }
 /* #249 D20: the viewer details' pair send and history buttons wrap inside the narrow rail at 200% zoom with wider fonts,
    where the toolkit's `nowrap` would push them past its edge. */
 .ca-tools #viewer-details > button { white-space: normal; /* D20 */ }
@@ -264,6 +276,9 @@ main:focus-visible { outline: 2px solid var(--ca-focus); outline-offset: 4px; }
 /* #249 D23: a parent with one preview, or none, fills the art row instead of leaving the concept's second cell empty. */
 .prompt-card .card-art > :only-child { grid-column: 1 / -1; /* D23 */ }
 .prompt-card .card-art > .small { display: grid; place-content: center; min-width: 0; color: var(--ca-muted); /* D23 */ }
+/* #437 D148: the card's selection checkbox is the concept's Images `.image-overlays`, placed over the art as on an image
+   card; `.media-art` carries the concept's `position: relative` there, the prompt card carries it here. */
+.prompt-card { position: relative; /* D148 */ }
 
 /* Prompt library shelf (#249 PR E). #249 D3: rows stay windowed, as the Images grid's do; the concept's `.library-grid`
    tracks and breakpoints lay them out, each row spans them, and appended rows never anchor the scroll. The row's