Eric HissandClaude Fable 5.1
4459af9ea6
Every node id, input name and existing widget position is unchanged, so saved
workflows load; new inputs are appended only.
Wildcards
- core/wildcards: Dynamic Prompts / Impact Pack grammar, tag queries, variables,
conditionals, concept / verbatim / loose bindings, reproducible seeds.
Wired into every enhancer node; reads ComfyUI/models/wildcards.
- New nodes: Wildcard Processor, Semantic Variation (direction axes),
Concept Survival Check. Starter lists in wildcards/.
- Random choices use a private generator; global random state is untouched.
Platform profiles
- core/profiles + profiles/*.yaml: one file per model (format, length,
negative prompt or none, quality tokens, text and dialogue syntax, vendor
rewriter prompt fetched once and cached). profiles_user/ overrides by id.
- New node: Platform Prompt Rewriter (prose, tags, edit instructions, JSON,
MiniMax H3 fields, Hailuo brackets), with validators for each format.
- Text-to-Image and Image-to-Image read profiles through platforms.py;
corrected facts for FLUX, Kontext, Chroma, Wan, AuraFlow, Kolors, NoobAI,
Qwen-Image, HunyuanImage. Negative prompt is empty for models without one.
- Hosted / API models: Seedream 5, Nano Banana, GPT Image, Midjourney V8,
FLUX.2 pro, Ideogram 3, Recraft V4, Seedance 2.0 / 2.5, Kling 3.0, Veo 3.1,
Gemini Omni Flash, Runway, Wan 3.0, Luma, hosted H3, Grok, Vidu, PixVerse.
Reports say how to switch the vendor's own prompt rewriting off.
- Illustrious: spaced tags, quality tags last, 25-80 words (OnomaAI paper).
Video nodes
- target_model and duration_seconds appended to the three video expanders.
Default target is MiniMax H3; wan22_video is the original path, unchanged.
A workflow saved earlier loads with the default until the input is set.
Also included: the Alice, Ideogram Magic Prompt and prompt-database nodes that
were already registered in __init__.py but had not been committed.
Fixes: v1 {animal} / {label:a|b} wildcards were inert in the video nodes;
Image-to-Video save_to_file raised NameError; T2I padded short-prompt models.
.gitattributes added (text=auto) so CRLF working files stop showing as
whole-file changes. prompt_database.db and the cached Ideogram prompt are
git-ignored.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019VpCi6nw5UECMdZ7ayYJvN
42 lines
2.7 KiB
YAML
42 lines
2.7 KiB
YAML
id: flux_kontex
|
|
name: "FLUX.1 Kontext [dev] (edit)"
|
|
developer: "Black Forest Labs"
|
|
summary: "Instruction-driven image EDITING model - not a long-prompt variant of FLUX, as this entry used to say. The prompt is an imperative edit plus a clause naming what must stay the same. (The id keeps its old spelling so saved workflows load.)"
|
|
modality: edit
|
|
tasks: [edit]
|
|
encoder: "T5-XXL + CLIP-L (512 tokens)"
|
|
encoder_family: t5
|
|
caption_style: instruction
|
|
validator: edit_instruction
|
|
languages: [en]
|
|
target_words: [12, 60]
|
|
max_tokens: 512
|
|
negative: ignored
|
|
quality_tokens: forbid
|
|
quoted_text: single
|
|
weights: false
|
|
structure: "[Change verb] [the specific thing, described not pronouned] to [exact new state], while maintaining [what must stay: facial features / composition / lighting / position, scale and pose]."
|
|
rules:
|
|
- "Precise verbs: Change, Replace, Add, Remove. 'Transform' can trigger a complete repaint - only use it together with an explicit preservation clause."
|
|
- "No pronouns: write 'the woman with short black hair', never 'she' or 'this person'."
|
|
- "Always end with a preservation clause: 'while maintaining the same facial features, composition and lighting' or 'keep everything else identical'."
|
|
- "Text edits: Replace 'original text' with 'new text' - quotes around both, case preserved; ask to keep the font style when wanted."
|
|
- "Style transfer: name the style, describe its traits (brushwork, palette), and state what to keep (composition, subject positions)."
|
|
- "Swapping a background: say 'keep the subject in the exact same position, scale, and pose'."
|
|
- "A big change is better as a sequence of small edits: write only the first step and say so."
|
|
avoid:
|
|
- "Describing the whole finished picture"
|
|
- "Vague verbs (make it better, improve)"
|
|
- "Several unrelated edits in one instruction"
|
|
legacy:
|
|
prompt_style: edit_focused
|
|
optimal_length: "concise (12-60 words)"
|
|
length_guidance: "One to three sentences: the change, then what stays the same."
|
|
detail_expectation: "Laser-focused on the requested change; name what must be preserved."
|
|
sources:
|
|
- {url: "https://docs.bfl.ai/guides/prompting_guide_kontext_i2i", supports: "official Kontext guide (could not be fetched directly on the check date)", confidence: medium}
|
|
- {url: "https://comfyui-wiki.com/en/tutorial/advanced/image/flux/flux-1-kontext", supports: "THIRD-PARTY mirror of the official guide: 512 tokens, transform and pronoun pitfalls, preservation phrasing, text quote syntax", confidence: medium}
|
|
- {url: "https://raw.githubusercontent.com/black-forest-labs/skills/master/skills/flux-image-best-practices/rules/flux1-models.md", supports: "official Kontext tips", confidence: high}
|
|
checked: "2026-09-20"
|
|
confidence: medium
|