Video Reader MCP
Evidence-first video MCP. Timeline document with ffprobe streams and subtitle cues.
People who work with developer tools, ffmpeg and local first and want it reachable from Claude, Cursor, VS Code, or another MCP client. The project is written in TypeScript.
NO LONGER MAINTAINED
LAST COMMIT 2026-09-25 · ★ 3 · VERIFIED 2026-09-30
MIT · TypeScript servers · how we verify → /methodology
01 · Install Video Reader MCP
Claude Code
claude mcp add sylphxai-video-reader-mcp -- npx -y @sylphx/video-reader-mcp Claude Desktop / Cursor / VS Code - add to config
{
"mcpServers": {
"sylphxai-video-reader-mcp": {
"command": "npx",
"args": [
"-y",
"@sylphx/video-reader-mcp"
]
}
}
} Same JSON for Cursor. For VS Code, rename the top-level key from `mcpServers` to `servers`.
Using another client? Same JSON, different key
Claude Desktop · mcpServers
Cursor · mcpServers
VS Code · servers
Windsurf · mcpServers
Zed · context_servers
Cline · mcpServers
Roo Code · mcpServers
Continue · mcpServers
LibreChat · mcpServers
Gemini CLI · mcpServers
Codex CLI · mcp_servers
Full setup guides: every client.
02 · Evidence
Security posture
What to check before giving this server access to your agent - from the registry, GitHub, and our own probes. We don't score safety; we show what's verifiable.
runs as local process (stdio) - runs on your machine with your user's permissions
repository archived on GitHub - read-only; no future fixes, security or otherwise
license MIT - declared in the repository
npm package @sylphx/video-reader-mcp - the @sylphx scope differs from the repo owner (sylphxai); confirm it's the project's official package
registry namespace io.github.SylphxAI is GitHub-verified and matches the repo owner
03 · What Video Reader MCP can do
Prose above is summarized from the project's README and registry record - no invented capabilities.
What you can build
An agent gets 3 documented tools across read & search, video, including read_video, search_video, video_evidence.
The 3 tools it gives your agent
Extracted from the project's README - what Video Reader MCP lets an agent do.
- read_video
- - Default fast profile: container metadata, streams, chapters, and embedded subtitles. Scenes are opt-in.
- search_video
- - Search embedded subtitle cues. It does not run speech recognition.
- video_evidence
- - Named follow-up: render, crop, or OCR one frame at a timestamp
Latest releases
v0.3.5 · 2026-09-23
Report the package version from the native binary. The multi-arch build compiled without CUE_PRODUCT_VERSION, so 0.3.4 answered 0.3.3.
v0.3.4 · 2026-09-23
The default read_video profile is fast: container metadata, streams, chapters, and embedded subtitles. Scenes, frames, OCR, and speech recognition stay off unless a caller asks for them, and the public docs say so.
v0.3.3 · 2026-09-23
0.3.1 · The MCP initialize response advertises the canonical product site, and the package gains discovery keywords. · serverInfo.websiteUrl still named the retired video-reader-mcp slug, so every connecting agent saw…
04 · Who maintains Video Reader MCP
Video Reader MCP is maintained by sylphxai. We track 13 MCP servers from sylphxai - 3 actively maintained, 2,932 combined GitHub stars, oldest repo from Apr 2025. Full record: all servers from sylphxai.
- PDF Reader MCP Evidence-first PDF MCP. Agent Document Twin with citeable page+bbox evidence. ★ 980
- Citra Citra - PDF answers with page-level proof: structure, tables, OCR, and page citations. ★ 963
- Locus Locus - the exact code chunk for the job: local BM25 search for agents. ★ 12
- Cue Cue - video answers with timestamp-level proof. Scenes and frames stay off until asked. ★ 3
- Image Reader MCP Evidence-first image MCP. Agent Media Twin with metadata and citeable OCR evidence. ★ 2
- anymd Any file → clean Markdown for AI agents: PDF, Office, EPUB, HTML, images, audio/video. Local MCP. ★ 963
05 · Facts
- repository
- github.com/sylphxai/video-reader-mcp
- category
- media - dead as of 2026-09-30.
- release cadence
- 7 releases in the last 90 days (latest 2026-09-23)
- registry
- io.github.SylphxAI/video-reader-mcp (deprecated, first published 2026-07-08)
- packages
- npm:@sylphx/video-reader-mcp
06 · Video Reader MCP FAQ
Is Video Reader MCP still maintained?
Its last commit was 2026-09-25 (as of 2026-09-30) - we classify it as dead.
What can Video Reader MCP do?
An agent gets 3 documented tools across read & search, video, including read_video, search_video, video_evidence.
How do I install Video Reader MCP?
Run `npx -y @sylphx/video-reader-mcp`. You can also paste the ready-made client config above.
Does Video Reader MCP run locally?
Yes - it's a stdio server: it runs on your machine (via npx) with your user's permissions. Your data stays local unless the server itself calls external APIs.
07 · Alternatives to Video Reader MCP
Alternatives to Video Reader MCP
Maintained media servers if Video Reader MCP isn't the fit.
- Openshorts Turn long videos into viral vertical shorts and publish them to TikTok, Instagram and YouTube. ★ 5,813 · 2026-09-28
- Claude Real Video Let any LLM watch a video locally - and search everything it has ever watched. ★ 2,191 · 2026-09-26
- Minutes The private, owned conversation-memory layer for AI. Record, transcribe, and search every meeting. ★ 1,518 · 2026-09-30
- Voicemode Natural voice conversations for AI assistants - STT/TTS via MCP ★ 1,383 · 2026-09-21
- Comfyui MCP MCP server + Claude Code plugin for ComfyUI: run workflows, generate images, manage models & VRAM. ★ 773 · 2026-09-28
- Nova3 D Structured, part-aware 3D generation for AI agents. Named-part GLB, preview URL, Blender script. ★ 726 · 2026-08-12