Use when the user asks to generate or edit images via the OpenAI Image API (for example: generate image, edit/inpaint/mask, background removal or replacement, transparent background, product shots, concept art, covers, or batch variants); run the bundled CLI (`scripts/image_gen.py`) and require `OPENAI_API_KEY` for live calls.
67
92%
Does it follow best practices?
Impact
65%
2.16xAverage score across 11 eval scenarios
Passed
No findings from the security scan
edit-invariants-specification
Uses edit command
0%
0%
Explicitly lists what to change
100%
0%
Explicitly lists invariants
100%
0%
batch-workflow-with-temp-cleanup
Creates JSONL in tmp/imagegen/
0%
0%
Correct JSONL format
0%
0%
Uses generate-batch command
0%
0%
Cleans up temporary JSONL
0%
0%
augmentation-without-invention
Adds layout constraints implied by use case
40%
0%
Does not invent new subjects
100%
100%
Stays focused on user request
100%
0%
restraint-keywords-to-avoid-tacky-output
Uses restraint keywords
100%
100%
Includes explicit Avoid: line
37%
100%
Avoids generic buzzwords
100%
100%
transparent-background-format-validation
Requests transparent background
0%
100%
Sets compatible output format
0%
100%
Uses edit or generate appropriately
0%
100%
verbatim-text-rendering
Quotes text verbatim
0%
0%
Specifies typography for main text
100%
0%
Specifies typography for subtitle
100%
0%
Requires exact rendering
0%
0%
structured-prompt-augmentation
Uses taxonomy classification
0%
100%
Structured prompt format
0%
85%
Includes constraints to avoid tacky output
40%
100%
Does not invent new creative elements
80%
100%
mask-based-inpainting
Uses edit command
0%
100%
Provides mask file
30%
100%
Provides input image
25%
100%
Specifies edit intent in prompt
0%
100%
quality-parameter-selection
Uses quality flag
0%
100%
Selects quality=high
0%
100%
input-fidelity-for-strict-edits
Uses edit command
0%
100%
Sets input_fidelity=high
0%
100%
Specifies identity preservation in prompt
100%
100%
cli-first-approach
Uses bundled CLI script
0%
100%
Does not create custom script
0%
100%
f627ab5
Table of Contents
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.