New Open Source Tool for Video Analysis with Claude
A new open source tool has been introduced that allows Claude to analyze videos from links, opening up new possibilities for AI users.
New Open Source Tool for Video Analysis with Claude
Now you don’t need to watch all of this — an open source skill has appeared that allows Claude to analyze videos from links almost like a human.
Tool Features
- Platform Support: Works with YouTube, Loom, TikTok, local files, and anything that supports
yt-dlp. - Analysis: The agent can analyze a competitor's launch by hooks, visuals, and structure, find bugs from screen recordings, create a brief summary of a long presentation with accurate timestamps, or extract only real changes from product videos without marketing.
Under the Hood
yt-dlpfirst retrieves available subtitles, andffmpegextracts frames considering scene changes to avoid wasting tokens on dozens of identical images. Claude receives frames with timestamps, and if there are no subtitles, Groq Whisper is connected.
Usage Instructions
For Claude Code:
/plugin marketplace add bradautomates/claude-video
/plugin install watch-video
For Codex, Cursor, and Gemini CLI:
npx skills add bradautomates/claude-video -g
Note: For videos longer than 10 minutes, it is better to specify the required segment using --start and --end, otherwise, a full analysis may significantly consume tokens.
Why it matters
AnalysisThis tool enables AI users to effectively analyze videos, which can significantly ease the process of extracting information from video content. It opens up new opportunities for research and analysis across various fields.
Discuss in community
Share your questions and insights with developers
Related in Ecosystem
Practical guides, core terminology, and agent skills
Hands-on Guide: Autonomous AI Agents on Open-Source Stack & Free APIs
If you want to build an AI agent, the first question is usually not \"which architecture to choose\", but \"where do I even start and what do I need\".
Part II. AI Agent Architecture: From Model to Memory and Tools
A deep engineering dive: tokenization, KV-cache, sampling determinism, the training stack, quantization, context engineering, the safety stack, production RAG, ReAct, and multi-agent swarm systems.
Tool Calling (Function Calling)
A low-level mechanism in language models that enables them to reliably generate validated parameters in JSON format for executing functions in external programming environments.
WebMCP & Browser-Native Tools
The adaptation of the Model Context Protocol (MCP) for direct operation within web browsers, enabling AI agents to interact with the DOM, LocalStorage, WebAssembly, and IndexedDB without local daemons.