
Crabcut
Connects your AI assistant to CrabCut's video clipping API for turning YouTube videos into vertical short-form content. Exposes five tools: generate_clips kicks off processing and returns a project ID, get_project_status polls for completion and returns scored clips with download URLs, list_projects shows your history, get_clip fetches individual clip details, and check_usage reports your credit balance. Clips come back reframed for 9:16 with auto-generated subtitles and a relevance score. Useful if you're building content workflows where an LLM identifies interesting videos and you want downloadable clips without manual editing. Supports webhooks for async workflows and works over standard MCP transports or their HTTP endpoint for n8n integrations.
- GitHub stars
- 1
- tools
- 5
- auth
- Secret needed
- transport
- STDIO
- env var
- 1
- last updated
- Mar 15, 2026
Tools
5 toolsPublic tool metadata for what this MCP can expose to an agent.
generate_clipsStart AI clip generation from a YouTube video. Returns {project_id, status, poll_url, estimated_minutes}. Poll with get_project_status every 10-15 seconds, or provide a callback_url for webhook notification.4 params
Start AI clip generation from a YouTube video. Returns {project_id, status, poll_url, estimated_minutes}. Poll with get_project_status every 10-15 seconds, or provide a callback_url for webhook notification.
callback_urlstringend_timenumberstart_timenumberurlstringget_project_statusReturns {id, name, status, step, error, expected_clips, duration, clips[]}. Clips sorted by score, each with clip_status (pending/exporting/completed/failed) and download_url when ready.1 params
Returns {id, name, status, step, error, expected_clips, duration, clips[]}. Clips sorted by score, each with clip_status (pending/exporting/completed/failed) and download_url when ready.
project_idstringlist_projectsReturns {projects[], total, limit, offset}. Each project has id, name, status, step, expected_clips, clips_count, duration, created_at.2 params
Returns {projects[], total, limit, offset}. Each project has id, name, status, step, expected_clips, clips_count, duration, created_at.
limitnumberstatusstringpending · processing · completed · completed_no_clips · failedget_clipReturns full clip details: id, project_id, title, duration, score, reason, export_status, export_quality, is_exported, video_url, thumbnail_url.1 params
Returns full clip details: id, project_id, title, duration, score, reason, export_status, export_quality, is_exported, video_url, thumbnail_url.
clip_idstringcheck_usageReturns {plan, credits_remaining, credits_total, period_start, period_end}. Check before generate_clips to confirm enough credits.
Returns {plan, credits_remaining, credits_total, period_start, period_end}. Check before generate_clips to confirm enough credits.
No parameter schema in public metadata yet.
Sponsors
Configuration
CRABCUT_API_KEY*secretYour Crabcut API key from app.crabcut.ai/developers
Readme
@crabcut/mcp-server
Turn YouTube videos into short clips — from Claude, Cursor, or any AI assistant that supports MCP.
Give it a YouTube link. It finds the best moments, reframes for vertical video (9:16), adds subtitles, and returns download-ready clips.
Quick Start
- Get an API key at app.crabcut.ai/developers
- Run:
CRABCUT_API_KEY=sk_live_... npx @crabcut/mcp-server
Setup
Claude Desktop / Cursor
Add to your config file — claude_desktop_config.json for Claude, .cursor/mcp.json for Cursor:
{
"mcpServers": {
"crabcut": {
"command": "npx",
"args": ["@crabcut/mcp-server"],
"env": {
"CRABCUT_API_KEY": "sk_live_..."
}
}
}
}
Cursor Directory
Smithery
Details
- Repository
- github.com/realcrabcut/crabcut-mcp-server
- Category
- Media & Entertainment
- Package
@crabcut/mcp-server- Transport
- STDIO
- Auth
- Required
- Registry
- MCP registry, active