MCP server that lets Claude visually inspect video files — extract frames, check overlays, karaoke subtitles and more
Wellknown found it in public sources; nobody has proven control of it yet. Claiming takes one click if the repository is under your GitHub account, or a small file on your domain otherwise. Verified owners get the badge, 15-minute checks, status alerts, edits that outrank crawled data, and a ranking boost.
Agents can do it too: POST https://wellknown.network/api/v1/claims with {"agent":"mcp-video-preview","method":"well_known_file"} — machine-readable steps at claim.json, guide at /docs/claim.
Everything here was measured by our prober or read from a registry. Nothing is self-reported.
Attributed to the source that supplied each field. Treated as claims, not facts.
# mcp-video-preview An MCP server that lets Claude visually inspect video files — extract frames, check text overlays, karaoke subtitles, background clips, and more. ## Requirements - Python 3.10+ - [ffmpeg](https://ffmpeg.org/download.html) installed and in your PATH ## Installation ```bash pip install mcp-video-preview ``` ## Setup Add to your Claude Code config (`~/.claude.json`): ```json { "mcpServers": { "video-preview": { "command": "mcp-video-preview", "args": [], "env": {} } } } ``` Or for **Claude Desktop** (`claude_desktop_config.json`): ```json { "mcpServers": { "video-preview": { "command": "mcp-video-preview" } } } ``` Restart Claude after editing the config. ## Tools ### `preview_video(path, n_frames=9)` Extracts N evenly-spaced frames from a video and returns them as images. Claude can then visually check text overlays, subtitles, background images, etc. ### `extract_frame(path, time_s)` Extracts a single frame at a specific timestamp (in seconds). ### `get_video_info(path)` Returns video metadata: duration, resolution, fps, file size. ## Example usage Once installed, just ask Claude: - *"Preview the video at C:/output/video.mp4"* - *"Check if the text overlay is readable in this video"* - *"Extract a frame at 10 seconds from this video"* ## License MIT
Mapped onto the structured taxonomy from declared text and observed tool names. Confidence shown for derived entries.
Every source is kept verbatim. Field changes are logged as events.