Works with Claude and ChatGPT

Your model goes where you work. Connect once, and your AI reads it before any work or career task.

What it does

  • Brief me: The AI reads your company's model and your strategy before it answers.
  • Tell the model: When your AI learns something new, it sends it to your inputs for you to accept.
  • Log a move: A leader can log a metric or a move update from the chat.
  • Nobody's words are logged: The app counts calls and tools, never what you said.

The three steps

  1. Connect. Paste one URL into your AI. Claude and ChatGPT sign you in; Claude Code and Cursor take a key the app makes.

  2. Tell it to use Expansion Playbooks. One standing instruction: "Before any work or career task, call Expansion Playbooks' start_here and follow its rules. When you learn something new, send it with propose."

  3. Test it. Click Test in the app, then ask your AI: "Brief me from Expansion Playbooks."

The four tools

  • Claude:Customize → Connectors → Add custom connector → paste the URL → sign in → Allow.
  • ChatGPT:chatgpt.com/plugins → Add custom MCP server → paste the URL as the public endpoint → sign in → Allow.
  • Claude Code:One terminal command the app writes for you, then a new session.
  • Cursor:One block for ~/.cursor/mcp.json the app writes for you, then restart Cursor.

What your AI can do with the model

list your companies · read the company model and the customer model · read the strategy and the moves · get a role or a playbook kit · propose something new to your inputs · log a metric or a move update (leaders).

Privacy

Your AI reads your model through your own sign-in. The app logs which tool was called and when, never a call's words.

Connect your AI