video-perception

1# Video Perception
2 
3You have access to video understanding tools via the claude-video-vision MCP server.
4 
5## Available Tools
6 
7- `video_analyze` — Analyze video structure with ffmpeg filters (scene changes, silence, motion, etc.). Use this BEFORE extracting frames to plan your strategy.
8- `video_watch` — Extract frames + process audio from a video. Supports variable FPS/resolution per segment.
9- `video_detail` — Drill into specific segments. Separates extraction from viewing — extract many frames, view few at a time.
10- `video_info` — Get video metadata without processing.
11- `video_configure` — Change settings (backend, resolution, enable_index, etc.).
12- `video_setup` — Check/install dependencies.
13 
14## Workflow
15 
16**IMPORTANT: You MUST follow these steps in order. Do NOT skip step 2.**
17 
181. Always start with `video_info` to get duration, resolution, and audio presence.
19   If the user gives a YouTube URL, pass the URL directly as `path`.
20   The MCP server downloads it with `yt-dlp`, prefers YouTube subtitles/auto-captions
21   for transcription, and falls back to the configured audio backend only when
22   captions are missing, empty, or suspiciously incomplete.
23 
242. **REQUIRED for videos > 30s:** Call `video_analyze` BEFORE extracting any frames.
25   This is NOT optional — it gives you structural data to make smart extraction decisions.
26   Select filters relevant to the user's question:
27 
28   | User intent | Filters to select |
29   |---|---|
30…