Connect Claude, ChatGPT, Claude Code, Cursor, or any MCP client to Pivot with the Pivot MCP server. Add one URL and sign in; no API key needed.
Pivot runs a Model Context Protocol (MCP) server. Connect an AI agent such as Claude, ChatGPT, Claude Code, or Cursor to it, and the agent can work in Pivot for you: read your spaces, rooms and messages, and send messages.
Add this URL to your agent, then sign in to Pivot in your browser when it asks. You don’t need an API key.
https://api.pivot.app/mcpYou can also copy it in Pivot: open Settings › Personal API keys. The Connect an AI agent section at the top shows the URL with a copy button.
Signing in works for both you and your organization, so it’s the right choice for Claude, ChatGPT, Claude Code, Cursor, and other ready-made agents. There’s no long-lived key to copy or store: the agent gets short-lived tokens that renew on their own.
On the authorize screen:
One authorization covers one identity, so authorize again to add another. The authorize screen works once and only for a few minutes; if it says the request has expired, start again from your agent.
Pivot, paste https://api.pivot.app/mcp, and select Continue.On Team and Enterprise plans, an Owner adds the connector in Organization settings › Connectors with the same steps. Each member then selects Connect under Customize › Connectors and signs in as themselves.
In ChatGPT on the web, open Plugins, select +, then Add custom MCP
server. Name it Pivot, enter https://api.pivot.app/mcp as the server URL,
and choose OAuth with dynamic client registration (DCR); Pivot doesn’t support
Client ID Metadata Documents. Select Create as a plugin, then sign in to
Pivot. Whether you can add custom MCP servers depends on your ChatGPT plan and
workspace settings.
claude mcp add --scope user --transport http pivot https://api.pivot.app/mcpThen run /mcp inside Claude Code, select pivot, and choose
Authenticate. --scope user makes Pivot available in all your projects.
Add Pivot to ~/.cursor/mcp.json (or .cursor/mcp.json in a project):
{ "mcpServers": { "pivot": { "url": "https://api.pivot.app/mcp" } }}Cursor then asks you to sign in to Pivot in your browser.
Any client that supports remote MCP servers (Streamable HTTP) with OAuth can connect with the URL alone. Pivot advertises its sign-in endpoints and accepts dynamic client registration, so there’s no client ID or secret to set up.
API keys are for code you run yourself, where no one is there to sign in: a script, a server, or an agent you built. Off-the-shelf clients such as Claude and ChatGPT should sign in instead.
Pass the key to your code in an environment variable, such as
PIVOT_API_KEY, and send it to https://api.pivot.app/mcp in the
Authorization header:
Authorization: Bearer <your Pivot API key>Treat a key like a password: keep it out of shared files and source control. See API key types for how the two kinds differ.
Agents connected as you are listed under Settings › Personal API keys › MCP authorizations, and agents connected as an organization under Org admin › Integrations. Revoke or delete one there to disconnect it.
Tool results pass through your agent host. Anything the agent reads from Pivot, such as messages, member names, and room details, is sent to the model provider behind that agent, under that provider’s terms. Only connect agents you are comfortable receiving that content.
There are no granular scopes yet. A connected agent can do everything the identity you authorized can do through the REST API, including sending messages and deleting content. Disconnect it if you stop trusting it.
Rate limits are the same as for the REST API, and each tool call counts as one request.
Was this guide helpful?