
MBQ Wedge	A ComfyUI custom node that sweeps any numeric parameter across a range — steps, CFG, guidance, denoise — queuing one image per value with a single click.
MBQ Browser	An OpenGL-accelerated desktop image viewer that reads ComfyUI's embedded prompt data from each PNG and overlays the swept parameter value directly on the canvas.

Together they form a tight review loop: sweep a parameter in ComfyUI → open the output folder in MBQ Browser → flip through the results with the swept value labelled on every image.

MBQ Wedge — ComfyUI node

Install

Copy (or symlink) the comfy_nodes/ folder into your ComfyUI custom nodes directory and restart ComfyUI:

ComfyUI/custom_nodes/ComfyUI-MBQWedge/

The node appears under MBQ → MBQ Wedge.

How it works

Add MBQ Wedge to any workflow and wire one output to any numeric input:

Output	Connect to
`int_value`	Any INT input (steps, width, …)
`float_value`	Any FLOAT input (CFG, guidance, denoise, …)

Set start, stop, and increment. Click Queue once — the MBQ JS extension intercepts the submission and expands it into N separate jobs automatically, one per sweep value.

Each output PNG has the exact swept value embedded in its prompt chunk, so MBQ Browser can read and display it without guessing — even if you rename or move the file.

Seeds advance correctly across sweep jobs (randomize / increment / decrement / fixed all work as configured).

Example — sweep CFG 3 → 7

parameter_name = "cfg"    (auto-filled when you connect the output)
start     = 3.0
stop      = 7.0
increment = 1.0

Wire float_value → KSampler cfg. Click Queue → 5 jobs, 5 PNGs, each labelled cfg: 3.00 through cfg: 7.00 in MBQ Browser.

Node display

will produce → N iterations — live count updates as you adjust the range
current → X — shows the first value; switches to integer display when wired to an INT input

MBQ Browser — desktop viewer

Install

Requirements: Python 3.11+, PySide6

pip install pyside6
python mbq_browser.py

python mbq_browser.py path/to/image.png   # open image + its folder
python mbq_browser.py path/to/folder/     # open folder

Features

Canvas

OpenGL-accelerated — smooth pan (left-drag), scroll-wheel zoom anchored at cursor
Middle-click-tap to reset zoom to 100%
Drag images in from Explorer or ComfyUI; drag thumbnails out to Explorer or ComfyUI

Zoom modes (mutually exclusive, each has a status bulb)

Key	Mode	Bulb
`Z`	Zoom Lock — preserve exact transform across image switches	Green
`F`	Fit Lock — auto-fit every image to the window	Teal
`R`	Reset — 100% zoom, clears both locks	—

Metadata panel

Three-tier display: (1) models, prompts, sampler params — (2) other scalars — (3) plumbing nodes (collapsible)
Colour-coded: cyan node headers · yellow keys · near-white values
Copy Summary — plain-text summary for notes or sharing
Copy Workflow — raw LiteGraph JSON for direct re-import into ComfyUI

MBQ Wedge integration

Wedge bulb (blue) lights when a swept image is detected
Swept parameter and value overlaid on the canvas: steps: 6
Overlay stays fixed in the corner — unaffected by zoom or pan

Keyboard shortcuts

Key	Action
`Ctrl+O`	Open image
`← / →`	Previous / next image
`Z`	Toggle zoom lock
`F`	Toggle fit lock
`R`	Reset zoom to 100%, clear locks
`S`	Toggle scroll freeze
`Ctrl+Q`	Quit

Status bulbs

Bulb	Colour	Meaning
Wedge	Blue	MBQWedge sweep detected in this image
Zoom	Green	Zoom Lock active
Fit	Teal	Fit Lock active
Scroll	Green	Scroll Freeze active
UnComfy	Amber	File may be a preview save, not a final output

File map

File	Role
`mbq_browser.py`	Main window — layout, menus, status bulbs
`mbq_logic.py`	Navigation, metadata display, filmstrip, wedge inference
`mbq_functions.py`	`ImageCanvas` (OpenGL view), `ImageFolder`, `WorkflowCache`
`mbq_parser.py`	PNG chunk extraction, ComfyUI prompt graph parsing
`comfy_nodes/mbq_wedge.py`	MBQWedge ComfyUI node
`comfy_nodes/js/mbq_wedge.js`	JS extension: widgets, queue intercept, seed handling
`dump_png_text_chunks.py`	CLI debug tool: dump raw PNG text chunks

Notes

ComfyUI only. MBQ reads the prompt chunk written by ComfyUI's SaveImage node. It does not support A1111, NovelAI, InvokeAI, or other tools.
Soft-fail. If a PNG has no prompt chunk, MBQ shows the image and leaves the metadata panel blank — it never crashes or shows garbage.
No ComfyUI runtime dependency — MBQ Browser reads files only.