ReddRadar
Agent skill
yt-transcribe
Use when user asks about YouTube video content, wants to know what a video says, needs information from a YouTube URL, or when video transcription would answer their question
Install this agent skill to your Project
npx add-skill https://github.com/mhagrelius/dotfiles/tree/main/.claude/skills/yt-transcribe
SKILL.md
YouTube Video Transcription
Overview
Use yt-transcribe to get the actual spoken content from YouTube videos. Web search and fetch tools cannot access video content - they only see metadata.
When to Use
Use this skill when:
- User shares a YouTube URL and asks what it says/contains
- User wants information that likely exists in a video
- User asks "what does this video talk about?"
- You need to extract spoken content from YouTube
Do NOT use:
- For video metadata only (title, description) - web search is faster
- When user explicitly wants just the link, not content
Quick Reference
| Flag | Purpose |
|---|---|
-q |
Required for LLMs - clean stdout, no progress noise |
-m tiny |
Faster transcription (less accurate) |
-m medium |
More accurate (slower, larger model) |
-o file |
Save to file instead of stdout |
-f srt/vtt |
Include timestamps |
Core Pattern
# Standard usage for LLM consumption
yt-transcribe "https://youtube.com/watch?v=VIDEO_ID" -q
# Faster but less accurate
yt-transcribe "https://youtube.com/watch?v=VIDEO_ID" -q -m tiny
# With timestamps
yt-transcribe "https://youtube.com/watch?v=VIDEO_ID" -q -f srt
Always use -q flag - suppresses progress output, gives clean transcript to stdout.
Workflow
- User provides YouTube URL or asks about video content
- Run:
yt-transcribe "URL" -q - Read and summarize the transcript for the user
- Answer their specific question using the transcript content
Common Mistakes
| Mistake | Fix |
|---|---|
| Trying WebFetch on YouTube URLs | YouTube blocks bots, use yt-transcribe |
Forgetting -q flag |
Progress output pollutes response, always use -q |
| Web searching for "what does video say" | Search finds metadata, not content - transcribe instead |
| Using without setup | Run yt-transcribe --setup first if binaries missing (exit code 5) |
Exit Codes
0- Success5- Binary not found (run--setup)3- Network error4- Transcription error
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
mhagrelius/dotfiles
building-tui-apps
Use when building interactive terminal dashboards or full-screen terminal applications; when implementing keyboard navigation, live data updates, or multi-panel layouts; when TUI is flickering, slow, or unresponsive; when handling terminal resize events
mhagrelius/dotfiles
building-mcp-servers
Use when building MCP servers in TypeScript, Python, or C#; when implementing tools, resources, or prompts; when configuring Streamable HTTP transport; when migrating from SSE; when adding OAuth authentication; when seeing MCP protocol errors
mhagrelius/dotfiles
deep-research
Use when user requests research requiring multiple sources, comprehensive analysis, or synthesis across topics - technical research, domain knowledge gathering, market analysis, or learning about complex subjects
mhagrelius/dotfiles
vercel-react-best-practices
React and Next.js performance optimization guidelines from Vercel Engineering. This skill should be used when writing, reviewing, or refactoring React/Next.js code to ensure optimal performance patterns. Triggers on tasks involving React components, Next.js pages, data fetching, bundle optimization, or performance improvements.
mhagrelius/dotfiles
designing-gnome-ui
Use when designing, implementing, or modifying UI for GNOME apps; before writing UI code; when reviewing existing UI for HIG compliance; when working with GTK 4/libadwaita or styling Qt/PySide6 for GNOME
mhagrelius/dotfiles
exa-search
Use when searching for concepts, ideas, or similar content without exact keywords; when user asks "find similar to...", needs semantic discovery, research across perspectives, or explicitly mentions exa
Didn't find tool you were looking for?