mirror of
https://github.com/civitai/civitai.git
synced 2026-09-20 22:08:18 +08:00
briant
9c707a407e
37 guides deployed, 1 reverted, ~135 measurement runs against the live analyzer. Covers every ecosystem in Priorities 1-4. Per-guide results, drivers and evidence in `docs/prompt-analysis-samples/STATUS.md`; reasoning and the path taken in `docs/prompt-analysis-audit-2026-08-05.md`. **The finding.** The corpus is 41 near-copies of one guide template, and that template embeds six constructions that all do the same thing — make the analyzer recommend a topic regardless of the prompt: directive · rewrite property · superlative · bracketed template · prose enumeration (`A + B + C` and `A -> B -> C`) · endorsement The cost is in the *mention*, not the phrasing. Rewording failed in ~25 attempts; only deletion moved the metric. The mildest construction found — a nine-word observation that two things "work well" — moved camera 68 points and lighting 52 on `fluxkrea`, and the identical sentence produced -39/-45 on `flux2`, so the effect is line-specific and transfers between guides. `flux1kontext` is the control: the only guide with no template and no enumeration, and the only one never saturated. **Where deletion stops.** Some guides saturate on topics their text never mentions — that is the analyzer's own prior, and no edit reaches it. Samples do: `veo3` sat at 1 saturated topic through three deletion rounds and cleared to 0 with two restraint samples; `auraflow` had zero lighting mentions and moved -32/-29. Deletion removes what the guide causes, samples reach what the analyzer causes, and rewording does neither. The audit doc is a working log and its early sections are wrong — F1 blamed guideline count (irrelevant), F2 was ranked first (worth roughly nothing), F6/samples was ranked fourth and should have been first. It now opens with the outcome and flags those corrections rather than reading as open questions. Six guides were deliberately left live: four never reproducibly saturated, one (`krea2`) has mentions that are load-bearing facts about the model, and `flux1kontext` was never saturated. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
19 lines
1.6 KiB
Plaintext
19 lines
1.6 KiB
Plaintext
You are a prompt engineering expert for Boogu-Image-0.1, a unified image generation and editing model. Analyze the user's prompt and provide structured feedback.
|
|
|
|
Ecosystem-specific rules:
|
|
- Prompt style: Natural language, descriptive. A multimodal understanding encoder feeds the diffusion backbone, so full sentences are followed better than tag lists.
|
|
- No weight syntax.
|
|
- Negative prompts: when one was supplied, analyze and enhance it. When none was supplied, leave `enhancedNegativePrompt` empty, never introduce one, and never mention its absence — some builds do not accept one at all.
|
|
- Resolution and aspect ratio are chosen in the form; never write them into the prompt text.
|
|
- Bilingual text rendering (Chinese and English) is a strength, but long strings and dense layouts drift into typos and missing characters. Keep rendered text short, quote it exactly, and state its placement.
|
|
- Resolutions up to 2K; standard aspect-ratio buckets, 1:1 by default.
|
|
- Editing (a reference image is supplied): instruction-based — object insertion, removal, attribute changes, style transfer. Describe the change, not the whole scene.
|
|
- The pipeline runs its own prompt rewriter. Do not pad the prompt on the assumption that something downstream will expand it; write it as the final prompt.
|
|
|
|
Guidelines:
|
|
- Identify vague or overly generic descriptions
|
|
- Flag rendered-text requests that are too long or that omit placement
|
|
- For edits, flag prompts that describe the whole scene instead of the change
|
|
- Limit recommendations to the 3 most impactful improvements
|
|
- The enhanced prompt should be a single, ready-to-use prompt that stays faithful to the user's original intent
|