Setting up MCP¶
Cohearence speaks the Model Context Protocol (MCP) — an open standard that lets AI clients like Cursor, Claude, and others connect to external tools and data. Once connected, your assistant can read your Cohearence data, write to it, and run real analysis on it, using everything documented in Ingesting Data and Exposing Analysis Over MCP.
There are two ways to authenticate a client:
| Personal Access Token | Google OAuth | |
|---|---|---|
| Best for | Cursor, Claude Code, and most other MCP clients that take a static URL + header | Claude Desktop / Claude.ai's "Add custom connector" flow |
| Setup | Copy a token from the app, paste it into your client's config | Sign in with Google right from the client — no copying anything |
| Revoking access | Revoke the token from Connect AI at any time | Disconnect the connector, or revoke access from your Google Account |
| Where it lives | A config file on your machine (or wherever your agent runs) | Nowhere you manage directly — the client holds a refresh token on your behalf |
Short version: if your client supports the full "Add custom connector" sign-in flow, use Google OAuth — it's one click and there's no secret to store. If it only takes a URL and a header (which is most agents today), use a Personal Access Token.
Continue to:
- Personal Access Token — generate a token and wire it into Cursor, Claude Code, Hermes, or OpenClaw.
- Google OAuth — connect Claude Desktop or Claude.ai without a token.