Skip to main content
Connect the GooseWorks MCP server to Claude, ChatGPT, Claude Code, Cursor, Codex, or another compatible client. Your agent can use your GooseWorks brands, campaigns, creatives, files, and tools from that client.

​
What is MCP?

MCP (Model Context Protocol) lets an AI app discover and call tools. The GooseWorks server gives your AI app access to your GooseWorks account. The CLI can configure that connection and install local skill instructions.

​
GooseWorks MCP server

  • URL: https://mcp.gooseworks.ai/mcp
  • Transport: Streamable HTTP
  • Authentication: OAuth where supported, or Authorization: Bearer <your-access-token>
Choose the setup that matches your client: Use credentials and server URLs from the same environment. A staging token should connect to the staging MCP server, not the production URL above.

​
Available tools

The Goose Growth tools reference is generated from the server source. Your client’s current tool list is the authority for what its deployed connection supports; a docs release and server release can arrive at different times. Older guides and clients may use names such as search_skills or submit_remix_batch. Compatibility aliases exist, but new integrations should use the canonical names advertised by their connection. See Ad Remixer MCP for creative generation, campaign planning, credit approval, and the Meta publishing boundary.

​
One-time setup with the CLI

Paste this into your agent — Claude Code, Cursor, Codex, or a chat app like Claude or ChatGPT:
Your agent reads that page and follows it. It asks you before it installs anything, then runs the GooseWorks CLI with --all (it sets up every coding agent on your machine at once), opens a browser for Google sign-in, checks the install, and carries on in the same chat: it sets up your brand there (it asks for your website and a couple of quick questions), then gets back to what you asked for. You don’t need to reopen your agent; the GooseWorks tools also show up in its next session. If your agent is a chat app like claude.ai, the Claude desktop app’s chat or ChatGPT, it gives you the MCP connector address instead. The page only ever runs npx gooseworks@latest …. It writes your agent’s skills folder, your agent’s MCP config (for Cursor that is the global Cursor MCP config, plus the project’s .cursor/mcp.json if the folder has one), and ~/.gooseworks/credentials.json. The MCP config entries contain your sign-in token. It writes no other files. Installing is free; some skills spend credits when they run. The check at the end also reports whether ffmpeg and a browser are present; those only matter for rendering video ads on your machine and do not block the install. For a manual terminal install:
Keep --mcp on agent-specific installs. --all already includes it. Restart your coding agent after setup so it loads the connection, then ask it to call account_whoami to confirm access.

​
Manual setup (without the CLI)

​
Step 1 — Create an MCP access token

  1. Open Settings → CLI & MCP and expand Manual setup.
  2. Click Create MCP access token. The token has the mcp:access scope and belongs to your user account.
  3. Copy it immediately. The complete token is shown only once.
The token authenticates your client. It does not bypass organization membership, credit requirements, or approvals. For paid generation, your account also needs a default workspace in the intended organization.

​
Step 2 — Add to your MCP client

For Cursor, use ~/.cursor/mcp.json. For Claude Code, use the MCP configuration in ~/.claude.json. Merge the server entry into any existing configuration rather than replacing your other servers.
For Codex, use the CLI setup above to configure its TOML settings. Clients that present a token field should receive the MCP access token in that field, following their connection guide.

​
Step 3 — Restart your editor

Restart your client, then ask GooseWorks which account is connected. It should call account_whoami. Check the organization before starting work.

​
Rotating or revoking

  • Rotate: create a new MCP access token, update each client using the old token, verify the connection, then revoke the old token.
  • Revoke: open Settings → CLI & MCP → Manual setup → Your MCP access tokens, click the token’s trash icon, and confirm. Clients using that token lose access.

​
Adding your own custom MCP servers

Connecting GooseWorks to an external AI app is different from connecting external services to your GooseWorks coworker. Manage the latter through Connections in the web app, using the integrations available to your account.