TeamGraph AI

Generate your API Key

To authorize your local MCP client to query the graph, you need a scoped API key.

Instructions: Sign in, navigate to API Keys in the dashboard, and create a key with the context.read and context.write scopes. Make sure to copy the raw key!

Install the MCP Package

Install the TeamGraph MCP CLI executable globally on your system.

npm install -g teamgraph

Add Server Config

Add TeamGraph to your agent client configurations.

Option A: Claude Desktop

Open your configuration file at:
~/Library/Application Support/Claude/claude_desktop_config.json

claude_desktop_config.json

{
  "mcpServers": {
    "teamgraph-live-brain": {
      "command": "teamgraph",
      "args": ["serve"],
      "env": {
        "TEAMGRAPH_SERVER_URL": "http://localhost:8000",
        "TEAMGRAPH_API_KEY": "tg_live_your_actual_api_key_here"
      }
    }
  }
}

Option B: Cursor

Navigate to Cursor Settings > Features > MCP.
Click + Add New MCP Server.
Set Name: TeamGraph, Type: command.
Command: teamgraph serve
Add Environment Variables:
- Name: TEAMGRAPH_SERVER_URL, Value: http://localhost:8000
- Name: TEAMGRAPH_API_KEY, Value: YOUR_API_KEY

Test that it works

Execute a connection test to confirm validation and connectivity from your terminal.

1. Run Login Authentication test

teamgraph login --api-key tg_live_xxx --server-url http://localhost:8000

2. Run Connection Status test

teamgraph status

If correctly connected, this returns a "validation": { "status": "valid" } confirmation output payload.