Skip to main content
Goose runs as an MCP server, so any AI app that speaks MCP can use your GooseWorks account directly — your brand, your creative, your data, billed to your credits. No new tab, no new habit. There are two ways to connect, and which one you use depends only on your app.

Paste a prompt

Claude Code, Cursor, Codex. Paste one prompt into your agent. It runs the install, signs you in, and sets up your brand in the same chat. No restart needed.

Sign in

Claude (claude.ai + Cowork), ChatGPT. Paste https://mcp.gooseworks.ai/mcp into the app’s connector settings and sign in. Nothing to install.

​
Which path is mine?

Claude appears twice on purpose. Claude Code is the terminal tool and takes the paste-a-prompt path. Claude the chat app (claude.ai, the desktop app, and Cowork) takes the sign-in path. They are different products with different setup — if you’re in a terminal, you want the prompt; if you’re in a browser or the Claude app, you want the connector.

​
Path A — paste a prompt (coding agents)

For Claude Code, Cursor, and Codex. Your agent runs the install, which signs you in with Google, registers the gooseworks MCP server in your app’s config, and installs the GooseWorks skills. It then checks the connection and sets up your brand in the same chat. 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. Prefer to type the command yourself? It’s in the CLI reference. You don’t need to restart your app. In the session that ran the install, the agent calls GooseWorks through the CLI; from the next session the GooseWorks MCP tools are listed directly. To check the connection yourself, ask: “Use the GooseWorks account_whoami tool to tell me which account is connected.” A successful tool response verifies access.
If the tool is unavailable, check that gooseworks is enabled in your client’s MCP settings. Rerun the install with your client flag and --mcp if needed, then restart the client. The CLI’s doctor command checks local video prerequisites and saved credentials; use account_whoami to verify the connection.

​
Requirements

  • Node 18+ (node --version). npx comes with it.
  • A browser on the same machine, for the Google sign-in. If none opens, the CLI prints the URL — paste it into any browser and finish there.

​
Path B — sign in (hosted connector)

For Claude (claude.ai, the desktop app, and Cowork) and ChatGPT on the web. Nothing to install and no token to copy: you add one URL and click Connect. The server address:
Add it as a custom connector in your app’s settings, click Connect, sign in with the same Google account you use for GooseWorks, and click Allow. Step-by-step with screenshots: Claude · ChatGPT (web) · ChatGPT work.

​
What you get, either way

It is the same server and the same account on both paths. Once connected, your AI app can, on your behalf:
  • Make, edit, and analyze ads — research a brand, turn templates into branded creative, re-roll and edit, and check how campaigns are doing.
  • Work with your files, chat history, and automations.
  • Query public web and social data — creator profiles, posts, competitor ads — with no provider keys of your own.
  • Find and run any GooseWorks skill on demand.
Full list: Goose Growth tools.

​
Next: actually make something

Connecting is step one. The five-minute walkthrough takes you from a fresh machine to three finished ads:

Quickstart — connect Goose to Claude and make three ads

The literal prompts to type and what comes back.