We only use cookies to enhance your experience. No cross-site tracking, no ads. Read our Cookie Policy.

Set up Supercutwith Codex.

Point Codex at Supercut’s MCP endpoint and bring recording context into your coding sessions.

01. Add the Supercut MCP server

Add the configuration below to ~/.codex/config.toml for every project, or to .codex/config.toml in a trusted project. The ChatGPT desktop app, Codex CLI, and Codex IDE extension share this configuration on the same host.

[mcp_servers.supercut]
url = "https://mcp.supercut.ai/mcp"

Add this to your Codex MCP configuration. The first time you use the server, Codex opens your browser to sign in to Supercut and authorize access.

02. Authorize Supercut

Start the OAuth flow from your terminal, then sign in, choose your Supercut workspace, and approve access in the browser.

codex mcp login supercut

03. Check the connection

List your configured servers from the terminal, or enter /mcp in the Codex terminal UI. Supercut should appear as enabled and authenticated.

codex mcp list

04. Try your first request

Start with a read-only request so you can see the recording context your client receives.

Find the Supercut walkthrough for this feature and compare what was requested with the current implementation.

Need client-specific troubleshooting? Read the Official OpenAI Codex MCP guide.

More setup guides...

MCP Overview

How Supercut MCP gives agents permission-aware access to recordings, transcripts, and frames.

Set up Supercut with Claude Code

Add the Supercut MCP server in Claude Code, then sign in only if Claude asks you to.

Set up Supercut with Claude Desktop

Connect Supercut as a remote connector in Claude Desktop and authorize access in the browser.

Set up Supercut with Cursor

Configure Cursor so its agents can search recordings and read transcripts from your editor.

Set up Supercut with VS Code

Add Supercut to VS Code for Copilot-powered workflows over your recordings.