{"slug": "let-claude-desktop-and-cursor-actually-watch-videos-mcp-fully-local", "title": "Let Claude Desktop and Cursor actually watch videos (MCP, fully local)", "summary": "A developer released claude-real-video, an open-source tool that lets AI agents like Claude Desktop and Cursor watch videos by extracting scene-aware keyframes and timestamped transcripts locally. The tool, available as an MCP server, reduces token usage by deduplicating frames and supports transcription via Whisper.", "body_md": "My AI agent learned to watch videos this week. Here is the 3-line version of how to give yours the same ability.\n\n[claude-real-video](https://github.com/HUANGCHIHHUNGLeo/claude-real-video) (MIT, 1.9k stars) extracts scene-aware, deduplicated keyframes plus a timestamped transcript from any video URL or local file — all processing on your machine. Since 0.8.0 it ships as an MCP server, so MCP clients can request a video directly.\n\n```\npip install 'claude-real-video[mcp]'\n```\n\nClaude Code:\n\n```\nclaude mcp add crv -- crv-mcp\n```\n\nClaude Desktop — add to `claude_desktop_config.json`\n\n:\n\n```\n{ \"mcpServers\": { \"crv\": { \"command\": \"crv-mcp\" } } }\n```\n\nCursor and any other MCP client: same stdio command, `crv-mcp`\n\n.\n\nTwo tools show up:\n\n`watch_video(source, max_frames, language, transcribe)`\n\n— downloads (URL) or reads (local file), extracts keyframes at scene changes, drops near-duplicates, transcribes with Whisper, and returns the transcript plus the first batch of frames as images straight into the conversation.`get_frames(source, start_index, count)`\n\n— pages through the rest of the frames.Analyses are cached per source under `~/.cache/crv-mcp`\n\n, so a follow-up question about the same video is instant.\n\nFixed-interval sampling wastes context on near-identical frames and still misses fast cuts. Scene detection keeps the frames that actually differ: a 58-second test clip goes from 58 sampled frames to 26 that matter. Fewer tokens, nothing missed. The benchmark folder in the repo has the full comparison, reproducible.\n\n`pip install 'claude-real-video[whisper]'`\n\nor `[fast]`\n\n); frames work without it.", "url": "https://wpnews.pro/news/let-claude-desktop-and-cursor-actually-watch-videos-mcp-fully-local", "canonical_source": "https://dev.to/huangchihhungleo/let-claude-desktop-and-cursor-actually-watch-videos-mcp-fully-local-2gnf", "published_at": "2026-08-03 02:16:29+00:00", "updated_at": "2026-08-03 02:39:05.226287+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-tools", "ai-agents", "computer-vision", "natural-language-processing"], "entities": ["claude-real-video", "Claude Desktop", "Cursor", "Whisper", "MCP"], "alternates": {"html": "https://wpnews.pro/news/let-claude-desktop-and-cursor-actually-watch-videos-mcp-fully-local", "markdown": "https://wpnews.pro/news/let-claude-desktop-and-cursor-actually-watch-videos-mcp-fully-local.md", "text": "https://wpnews.pro/news/let-claude-desktop-and-cursor-actually-watch-videos-mcp-fully-local.txt", "jsonld": "https://wpnews.pro/news/let-claude-desktop-and-cursor-actually-watch-videos-mcp-fully-local.jsonld"}}