القائمة
Self-evolving skills. Record user feedback, update methodology and rules. Triggers: update rules, record feedback, improve skill, 更新规则, 记录反馈, 改进skill
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
End-to-end QCut CLI workflow for cleaning talking-head and vlog footage, applying restrained portrait filters and skin smoothing, optionally cutting out the speaker and replacing the background, regenerating time-aligned SRT captions after the cuts, and exporting both an editable MP4 plus SRT and a hard-captioned MP4. Use for vlog editing, talking-head cleanup, portrait enhancement, background replacement, 口播剪辑, 剪口播, 去口头词, 去停顿, 人像滤镜, 美颜, 抠像换背景, 自动字幕, 中文字幕, or exporting a captioned social video.
Unified QCut media toolkit — organize project files, process media with FFmpeg, generate AI content, control the QCut editor with native CLI commands, generate video prompts, and test MCP preview. Use when the user asks about any media workflow, file organization, video processing, AI generation, editor control, video prompts, or content pipeline task.
Unified QCut media toolkit — organize project files, process media with FFmpeg, generate AI content, control the QCut editor with native CLI commands, generate video prompts, and test MCP preview. Use when the user asks about any media workflow, file organization, video processing, AI generation, editor control, video prompts, or content pipeline task.
Run QCut's native TypeScript pipeline CLI for AI content generation, video analysis, transcription, YAML pipelines, ViMax agentic video production, and project management. Also use for editor HTTP automation tasks (state snapshots, events, transactions, and notification bridge control) when user needs deterministic state-aware control.
agent-im 会话技能 - 通过 liblib.tv 的 AI 能力生成和编辑图片/视频。覆盖场景包括：生成（文生图、文生视频、图生视频、做动画、画一个xxx、来段xxx）、编辑修改（把xxx换成yyy、去掉xxx、加上xxx、改成xxx、调整xxx、局部修改、改镜头）、风格转换（风格迁移、转绘、换风格）、视频续写延长、复刻视频/TVC/宣传片、短剧/短漫剧生成、音乐MV生成、产品广告/展示片制作、分镜/故事板设计、教育视频/短视频制作。当用户提到 liblib、libtv、上传参考图/视频、查看生成进度时也应触发。关键判断：只要用户的请求涉及 AI 图片或视频的创作、生成、编辑、修改，无论措辞如何（如"画只猫"、"做个海报"、"把纸船换成爱心"、"这个视频帮我改一下"、"帮我复刻这段视频"、"用这首歌做个MV"、"一句话生成短剧"），都必须触发此技能。
Export, preprocess, and fix GitHub PR review comments. Use when user wants to export PR comments, evaluate code reviews, or fix review feedback from CodeRabbit/Gemini/Devin bots.
Let the Agent learn from mistakes and continuously improve
User: Record the issue we just had
User: Update the speech error detection rules
User: Remember this lesson
User: 记录一下刚才的问题
User: 更新口误识别的规则
| Content Type | Target File | Example |
|---|---|---|
| User profile | CLAUDE.md | Preferences, habits |
| Methodology + feedback | */tips/*.md | Rules, lessons |
User triggers ("that just failed", "record this")
↓
[Auto] Trace back context, find the problem
↓
[Auto] Read target file fully, understand structure
↓
[Auto] Integrate into the appropriate section (not just append!)
↓
[Auto] Feedback log only records events, doesn't repeat rules
↓
Report update results
Key: Don't ask "what's the problem" — analyze from context directly!
## Feedback Log
### 2026-01-14
- Lesson: Must generate deletion task list at end of review
- Lesson: Confirm errors and silence separately
Only adding to feedback log = rules scattered at the bottom, will repeat mistakes.
- Review marked silence, but missed during cut## IV. Review Format
(Add deletion task list template)
## V. Confirmation & Execution Flow ← Add this section if missing
(Add separate confirmation for errors and silence)
## Feedback Log
### 2026-01-14
- Review marked silence, but missed during cut (only cut errors)
Bad:
User: That just failed, update the skills
AI: What problem did you find? ← Should NOT ask!
Good:
AI: [Auto trace back context, find failure point]
AI: [Execute update]
Bad:
AI: Updated. Added 3 lessons to feedback log ← Only appended!
Good:
AI: [Read full file, understand structure]
AI: [Integrate into appropriate sections]
AI: [Feedback log only records events]
AI: Updated: Added Chapter V "Confirmation Flow", updated Chapter IV template
Principle: Rules must be integrated into the body; feedback log is just an event journal.