원클릭으로
evaluation-and-repair
Visual and content evaluation prompts, metric catalogs, and repair memory.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
Visual and content evaluation prompts, metric catalogs, and repair memory.
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
Plan page-level PPT outlines, audience adaptation, and visual-mode semantics for the deck.
PPT visual production rules, local enhancements, and vendor skill composition for theme, assets, and slide generation.
Extract text and tables from Word (.docx) files. Use when a user uploads a Word document that needs to be processed.
Extract text and tables from Markdown (.md) files. Use when a user uploads a Markdown document that needs to be processed.
Extract text and tables from PDF files. Use when a user uploads a PDF document that needs to be processed.
Extract text and tables from PowerPoint (.pptx) files. Use when a user uploads a PPTX presentation that needs to be processed.
| name | evaluation-and-repair |
| description | Visual and content evaluation prompts, metric catalogs, and repair memory. |
| phase | evaluation-and-repair |
| order | 50 |
| default_prompt_mode | static |
| asset_groups | [{"kind":"skill_doc","pattern":"SKILL.md","purpose":"Phase overview for content evaluation, visual evaluation, and repair.","injection_points":["docs:evaluation-and-repair"],"tags":["documentation"]},{"kind":"reference","pattern":"*_system.md","purpose":"System prompt reference for evaluation and repair sub-stages.","injection_points":["content_evaluation:system","visual_evaluation:system","deck_coherence_review:system"],"tags":["system-prompt"]},{"kind":"reference","pattern":"*policy.json","purpose":"Machine-readable evaluation and repair policies.","injection_points":["runtime:evaluation-policy"],"tags":["policy"]},{"kind":"reference","pattern":"*aliases.json","purpose":"Static metric alias catalogs for evaluation normalization.","injection_points":["runtime:catalog"],"tags":["catalog"]},{"kind":"reference","pattern":"default_principle_descriptions.json","purpose":"Fallback principle descriptions for content evaluation.","injection_points":["content_evaluation:user-fragment"],"tags":["catalog"]},{"kind":"reference","pattern":"repair_instruction_map.json","purpose":"Error-signature to repair-instruction mapping for evaluation retries.","injection_points":["evaluation:repair-map"],"tags":["repair"]},{"kind":"reference","pattern":"repair_patterns.md","purpose":"Human-authored repair heuristics for evaluation failures.","injection_points":["docs:repair-heuristics"],"tags":["repair"]},{"kind":"template","pattern":"content_evaluation_user.txt","purpose":"Main user prompt template for content evaluation.","injection_points":["content_evaluation:user"],"tags":["user-prompt"]},{"kind":"template","pattern":"visual_evaluation_user.txt","purpose":"Main user prompt template for visual evaluation.","injection_points":["visual_evaluation:user"],"tags":["user-prompt"]},{"kind":"template","pattern":"deck_coherence_review_user.txt","purpose":"Main user prompt template for deck-level coherence review.","injection_points":["deck_coherence_review:user"],"tags":["user-prompt"]},{"kind":"template","pattern":"*retry*.txt","purpose":"Retry feedback and retry wrappers for evaluation failures.","injection_points":["content_evaluation:repair-feedback","visual_evaluation:repair-feedback","deck_coherence_review:repair-feedback"],"tags":["repair-feedback"]},{"kind":"template","pattern":"*metric*.txt","purpose":"Metric rendering fragments for content evaluation prompts.","injection_points":["content_evaluation:user-fragment"],"tags":["metrics"]},{"kind":"template","pattern":"*progress*.txt","purpose":"Progress-reporting fragments for streaming evaluation responses.","injection_points":["content_evaluation:streaming"],"tags":["streaming"]},{"kind":"template","pattern":"*default*.txt","purpose":"Fallback reason and suggestion fragments for evaluation normalization.","injection_points":["content_evaluation:normalization"],"tags":["fallback"]},{"kind":"template","pattern":"visual_evaluation_failed_*.txt","purpose":"Canonical failure descriptions and suggestions for visual evaluation retries.","injection_points":["visual_evaluation:repair-feedback"],"tags":["repair"]}] |
This skill governs visual QA, content QA, and repair-oriented evaluation.