{"$schema":"https://wellknown.network/schemas/agent-record-v1.json","schemaVersion":"1","id":"ag_e2c43nqj32y6","handle":"video-context-mcp","url":"https://wellknown.network/agents/video-context-mcp","links":{"self":"https://wellknown.network/agents/video-context-mcp/record.json","html":"https://wellknown.network/agents/video-context-mcp","markdown":"https://wellknown.network/agents/video-context-mcp/record.md","api":"https://wellknown.network/api/v1/agents/video-context-mcp","status":"https://wellknown.network/api/v1/agents/video-context-mcp/status","claim":"https://wellknown.network/agents/video-context-mcp/claim","claimApi":"https://wellknown.network/api/v1/claims","claimDescriptor":"https://wellknown.network/agents/video-context-mcp/claim.json","badge":"https://wellknown.network/agents/video-context-mcp/badge.svg","openapi":"https://wellknown.network/openapi.json"},"ard":{"identifier":"urn:air::server:video-context-mcp","type":"application/mcp-server-card+json"},"kind":"mcp_server","declared":{"name":"video-context-mcp","summary":"Local video and animated-GIF RAG for MCP-compatible coding agents","description":"# Keyframe\n\n**Local video and animated-GIF RAG for coding agents.** Keyframe turns a tutorial,\nscreen recording, or animation into timestamped transcript segments, searchable on-screen text,\nrepresentative frames, and reconstructed code. Ask what was *said*, what was\n*shown*, or both—and inspect the source frame before trusting uncertain OCR.\nIt runs locally in Codex, ChatGPT desktop, Claude Code, Cursor, and Google\nAntigravity/Agy through the same MCP server.\n\nKeyframe is deliberately split into two parts:\n\n- a local MCP server performs deterministic acquisition, OCR, indexing, and\n  retrieval; and\n- a small workflow skill teaches any connected agent to retrieve narrowly,\n  verify visual evidence, and cite timestamps.\n\nThe server does not call an LLM. In the Build Week workflow, Codex running\nGPT-5.6 reasons over Keyframe's evidence, changes code, and runs the tests.\n\n### Topic discovery versus video analysis\n\nAn explicit Keyframe invocation selects the analysis capability; it does not\nchange the subject of the question. When a user asks for videos about a topic\nwithout supplying a source, the connected agent uses its normal web-search\ncapability to identify up to three individual public videos. If one centrally\nmatches the requested subject and task, Keyframe ingests that one source and\nadds timestamped evidence. Adjacent or keyword-only results are recommended\nwith a caveat rather than ingested. Keyframe itself does not search the public\nweb, and its library-wide `video_search` covers only previously indexed media.\n\n![Conceptual Keyframe workflow: video moments become said/shown evidence, verified code, and passing tests](https://raw.githubusercontent.com/MatthewOscar/Keyframe/main/docs/design/keyframe-devpost-hero.png)\n\n*Product-story concept; Keyframe is a local MCP server and plugin, not a hosted UI.*\n\n## Quick start\n\n### Prerequisites\n\nKeyframe v0.3.1 supports CPython 3.12, 3.13, and 3.14. Install these native\ntools before starting:\n\n- FFmpeg and `ffprobe` f…","publisher":{"name":"Matthew Wyatt","url":null},"homepage":"https://github.com/MatthewOscar/Keyframe","repository":"https://github.com/MatthewOscar/Keyframe","version":"0.3.1","license":null,"protocols":["mcp"],"tags":["animated-gif","claude-code","codex","cursor","mcp","ocr","rag","video"],"pricing":null,"endpoints":[{"url":"pypi:video-context-mcp","type":"package_pypi","auth":null,"probeable":false}],"skills":null,"tools":null,"extra":null,"attribution":{"kind":"pypi","name":"pypi","repoUrl":"pypi","summary":"pypi","version":"pypi","description":"pypi","homepageUrl":"pypi","publisherName":"pypi"}},"derived":{"capabilities":[{"slug":"documents.ocr","name":"OCR","confidence":1,"provenance":"declared"},{"slug":"data.vector-search","name":"Vector Search","confidence":1,"provenance":"declared"}],"categories":["data","documents"],"language":"en"},"observed":{"status":"unknown","statusReason":"Distributed as a package to run locally; no network endpoint to check.","lastOkAt":null,"lastProbedAt":null,"statusComputedAt":null,"reliability30d":null,"latestObservations":[],"tools":null,"package":{"name":"video-context-mcp","registry":"pypi","observedAt":"2026-09-10T15:25:11.471Z","publishedAt":"2026-07-21T21:01:51.500805Z","latestVersion":"0.3.1"}},"verification":{"claimed":false,"claimedAt":null,"proofs":[]},"provenance":{"sources":[{"source":"pypi","key":"video-context-mcp","url":"https://pypi.org/project/video-context-mcp/","firstSeenAt":"2026-09-10T15:23:32.683Z","fetchedAt":"2026-09-10T15:23:32.683Z","normalizedAt":"2026-09-10T15:23:32.683Z"}]},"firstSeenAt":"2026-09-10T15:23:32.683Z","updatedAt":"2026-09-10T15:25:11.471Z"}