Video-Learner
Declared Requirements
This skill requires the following tools and access:
- - Tools: python3, yt-dlp, ffmpeg, whisper (Python package), node
- Dependency: douyin-download (requires node)
- Write path: ~/.openclaw/workspace/skills/
Trigger
When the user sends a Douyin/BiliBili/YouTube video link, this skill is invoked.
Processing Flow
- 1. Create temp directory in
/tmp/ for video download - Download video using yt-dlp or douyin-download
- Extract audio using ffmpeg
- Transcribe audio to text using Whisper (local)
- Analyze text content using the agent's LLM capability
- Display analysis results to user
- After user confirmation, generate SKILL.md to ~/.openclaw/workspace/skills//
- Delete temp video files after processing
Security Notes
- - Whisper: Runs locally, no audio sent to external services
- Content analysis: Uses the agent's own model capability
- Temp files: Automatically cleaned up after processing
- User confirmation: Generated new Skill requires explicit user approval
- Generated Skills are stored in ~/.openclaw/workspace/skills/, can be deleted anytime
Limitations
- - Only processes user-provided video links
- Does not proactively fetch other content
- Does not store video files
视频学习器
声明需求
该技能需要以下工具和访问权限：
- - 工具：python3、yt-dlp、ffmpeg、whisper（Python包）、node
- 依赖项：douyin-download（需要node）
- 写入路径：~/.openclaw/workspace/skills/
触发条件
当用户发送抖音/哔哩哔哩/YouTube视频链接时，该技能被调用。
处理流程
- 1. 在/tmp/中创建临时目录用于视频下载
- 使用yt-dlp或douyin-download下载视频
- 使用ffmpeg提取音频
- 使用Whisper（本地）将音频转录为文本
- 使用代理的LLM能力分析文本内容
- 向用户显示分析结果
- 在用户确认后，生成SKILL.md到~/.openclaw/workspace/skills/<新技能名称>/
- 处理完成后删除临时视频文件
安全说明
- - Whisper：本地运行，不向外部服务发送音频
- 内容分析：使用代理自身的模型能力
- 临时文件：处理完成后自动清理
- 用户确认：生成的新技能需要用户明确批准
- 生成的技能存储在~/.openclaw/workspace/skills/中，可随时删除
限制
- - 仅处理用户提供的视频链接
- 不会主动获取其他内容
- 不存储视频文件
