메뉴
When the user wants to create or edit images → generate with Seedream (text-to-image + image-to-image).
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
When prompt, tool, or agent logic changes need regression testing → run eval benchmarks to detect regressions or improvements.
When a task can be parallelized (bulk code review, multi-module refactor, batch analysis) → split into subtasks running in isolated git worktrees via multiple workers.
When the user wants to research academic papers → search ArXiv, extract PDFs, summarize/review/Q&A, generate experiment code.
When you need to interact with web pages using the user's logged-in browser → control Chrome via @playwright/cli daemon, reusing cookies/session with persistent tabs.
When user needs AnyGenIO content generation → unified CLI for task execution with progressive disclosure.
When you need to persist files (reports, docs, evidence) beyond the session → create/query/delete durable artifacts.
| name | image-creation |
| description | When the user wants to create or edit images → generate with Seedream (text-to-image + image-to-image). |
| triggers | {"intent_patterns":["生成图片|画|draw|image|图片|插图|illustration|设计图|海报","帮我画.*|画一个|画个|generate.*image|create.*image","logo|图标设计|icon.*design|封面|cover.*image|thumbnail","壁纸|wallpaper|背景图|banner|头像|avatar|profile.*pic","效果图|mockup|示意图|概念图|concept.*art","修改.*图片|edit.*image|换个.*风格|style.*transfer|重新生成","图生图|image.*to.*image|以图生图|参考.*这张"],"context_signals":{"keywords":["图片","image","draw","画","生成","设计","logo","海报","封面","banner","头像","壁纸","概念图","mockup"]},"confidence_threshold":0.6} |
| priority | 7 |
| exclusive_group | image |
| requires_tools | ["bash","write"] |
| max_tokens | 200 |
| cooldown | 30 |
| output | {"format":"markdown","artifacts":true,"artifact_type":"image"} |
Generate images via Seedream.
ARK_API_KEY (required)SEEDREAM_TEXT_ENDPOINT_ID (optional; fallback: SEEDREAM_TEXT_MODEL -> built-in default model)SEEDREAM_I2I_ENDPOINT_ID (required for refine)1920*1920. Smaller inputs (for example 1024x1024) are auto-upscaled.success=true only when the output file is actually written and non-empty.b64_json or url; otherwise the call fails./tmp unless output is provided.watermark defaults to false (no "AI generated" watermark). Set to true only when you explicitly need watermark.# Text to image
python3 skills/image-creation/run.py generate --prompt 'white cat in moonlight' --style realistic --watermark false
# Image to image
python3 skills/image-creation/run.py refine --image_path /tmp/cat.png --prompt 'add starry sky background' --watermark false
| name | type | required | notes |
|---|---|---|---|
| prompt | string | yes | image description |
| style | string | no | style tag (default: realistic) |
| size | string | no | WIDTHxHEIGHT, default 1920x1920 |
| watermark | bool | no | default false; whether to enable API watermark |
| output | string | no | output file path (default /tmp/seedream_<ts>.png) |
| name | type | required | notes |
|---|---|---|---|
| image_path | string | yes | input image path |
| prompt | string | yes | refinement instruction |
| watermark | bool | no | default false; whether to enable API watermark |
| output | string | no | output path (default /tmp/seedream_refined_<ts>.png) |