CutPro MCP

A Model Context Protocol (MCP) server that turns long videos into viral clips with AI. It exposes the full CutPro API as tools, so an LLM can run the whole flow: analyze a video, clip the best moments, render the final MP4, and publish to TikTok, Instagram and YouTube.

Key features

End to end. All 34 v1 endpoints as tools: workspace, balance, videos, clipping, clips, templates, renders, posts and connections.
Token efficient. Results are compact and projected to the fields that matter; list_clips is rating sorted, capped, and omits long signed URLs unless asked.
Runs everywhere. stdio for local clients (Claude Code, Cursor, Claude Desktop, Windsurf, VS Code, Cline, Zed) and a hosted Streamable HTTP endpoint with OAuth for ChatGPT and Claude.ai.

Getting started

Requirements

Node.js 18 or newer.
A CutPro account on the Pro plan and an API key. Generate one at cut.pro/studio/me/api-keys.
An MCP-compatible client.

Standard config

Most clients use the same JSON. Add your API key under env:

{
  "mcpServers": {
    "cutpro": {
      "command": "npx",
      "args": ["-y", "@cutpro/mcp"],
      "env": { "CUTPRO_API_KEY": "<your-api-key>" }
    }
  }
}

Install

After installing via a button, add your CUTPRO_API_KEY to the server's env.

Claude Code

claude mcp add cutpro --env CUTPRO_API_KEY=<your-api-key> -- npx -y @cutpro/mcp

Claude Desktop

Add to claude_desktop_config.json (Settings, Developer, Edit Config):

{
  "mcpServers": {
    "cutpro": {
      "command": "npx",
      "args": ["-y", "@cutpro/mcp"],
      "env": { "CUTPRO_API_KEY": "<your-api-key>" }
    }
  }
}

Cursor / Windsurf / VS Code (manual)

Add the standard config above to the client's MCP settings (mcp.json / mcpServers).

Cline

Open the MCP Servers panel, choose Configure, and add the standard config above.

Gemini CLI

gemini mcp add cutpro npx -y @cutpro/mcp -e CUTPRO_API_KEY=<your-api-key>

Codex

Add to ~/.codex/config.toml:

[mcp_servers.cutpro]
command = "npx"
args = ["-y", "@cutpro/mcp"]
env = { "CUTPRO_API_KEY" = "<your-api-key>" }

ChatGPT and Claude.ai (hosted, no install)

Use the hosted server. Add a custom connector pointing to:

https://mcp.cut.pro

You authorize with your CutPro API key on a consent page (OAuth), so no local setup is needed.

Configuration

The server is configured with environment variables.

Variable	Description	Required
`CUTPRO_API_KEY`	Your CutPro API key (Pro plan).	Yes (stdio)
`CUTPRO_WORKSPACE_ID`	Selects the workspace for multi-workspace keys.	No
`CUTPRO_API_URL`	Override the API base URL. Defaults to `https://api.cut.pro/api/v1`.	No

Self-hosting the remote (Streamable HTTP + OAuth)

Variable	Description
`MCP_TRANSPORT=http` / `PORT`	Serve Streamable HTTP at the root instead of stdio.
`MCP_OAUTH=1`	Enable the full OAuth 2.1 layer (discovery, DCR, PKCE) for browser clients.
`MCP_PUBLIC_URL`	Public endpoint, e.g. `https://mcp.cut.pro`. Its origin becomes the OAuth issuer.
`MCP_REDIS_URL`	Back OAuth state with Redis so it survives restarts and scales across instances.

MCP_TRANSPORT=http PORT=8787 MCP_OAUTH=1 \
MCP_PUBLIC_URL=https://mcp.cut.pro MCP_REDIS_URL=redis://127.0.0.1:6379 \
npx -y @cutpro/mcp

In OAuth mode the user authorizes with their own API key on a consent page; the access token maps server side to that key. Without MCP_REDIS_URL, an in-memory store is used (single instance, state lost on restart).

Tools

Workspace and balance

get_workspace: the workspace this key resolved to, with plan and role.
get_balance: current credit balance.
get_balance_history: ledger of credits added and consumed.

Videos and uploads

list_videos: your source video library.
start_upload: get a presigned URL to upload your own file (max 2 GB; .mp4/.mov/.webm/.mkv).
complete_upload: register a finished upload and get its credit cost.
delete_video: delete a source video and its submissions.

Clipping

analyze_video: preview metadata and credit cost of a public URL (free).
submit_clipping: start AI clipping (charges credits).
list_submissions: clipping jobs for a video.
get_submission: poll a submission until completed or failed.
delete_submission: delete a submission and its clips.

Clips and templates

list_clips: clips of a completed submission, rating sorted (URLs opt-in).
apply_template: apply an editing template to clips in bulk.
delete_clip: delete a single clip.
list_templates: editing templates to apply to clips.

Renders

render_clip: render a clip to a final MP4.
get_render: poll a render until completed.
get_render_download: signed download URL of a completed render.
cancel_render: cancel or delete a render.
get_render_limits: render quota for the workspace.
start_bulk_download / get_bulk_download: bundle several renders into one download.

Posts and connections

create_post: publish rendered clips to connected accounts (immediate or scheduled).
list_posts / get_post / update_post / delete_post: manage posts.
publish_post: trigger publishing now.
retry_post_item / delete_post_item: handle individual targets.
list_connections / get_connection: connected social accounts.

Each tool carries read-only / write / destructive annotations so clients can plan calls.

License

MIT