# Use Engagaloo with your AI

> Connect Claude, ChatGPT, Grok or any MCP assistant straight to Engagaloo: its own MCP address, sign-in or API key, 1 tool and 58 actions.

Source: https://engagaloo.tech/docs/use-with-your-ai/

Engagaloo runs its own MCP server. Claude, ChatGPT, Grok or any assistant that supports MCP connects straight to Engagaloo, signs in as you and works in your own Engagaloo account. You need nothing else.

## The Engagaloo address

```text title="The MCP address of Engagaloo"
https://app.engagaloo.tech/mcp
```

The server speaks MCP over HTTP and answers `POST /mcp` on the Engagaloo address. It takes a sign-in (OAuth) or an API key.

## Connect your AI

### Claude

1. In Claude, open Settings, then Connectors, then Add custom connector.
2. Paste the Engagaloo address and save.
3. Sign in with your Engagaloo account when Claude asks (OAuth: Claude registers itself and you approve it once).

### ChatGPT

1. In ChatGPT, open Settings, then Connectors, and add a remote MCP server.
2. Paste the Engagaloo address.
3. Sign in with your Engagaloo account when it asks.

### Grok and other assistants

Any assistant that supports remote MCP servers works the same way: add a remote MCP server, paste the Engagaloo address and sign in when it asks.

### Scripts and clients that cannot sign in

1. In Engagaloo, make an API key on its API keys page. It is shown once, so copy it then.
2. Point your client at the Engagaloo address.
3. Send the key as a bearer token.

::mcp-config

> [!WARNING]
> Never paste a key into a chat. Use a key only in the configuration of a client that cannot sign in, name each key after what uses it, and revoke it in Engagaloo when you stop using that client.

## What your AI can use

Engagaloo has 1 tool and 58 actions in all. Every tool also answers `help`. Ask a tool for its actions with `action: "help"`, and for the arguments of one action with `action: "help"` and `of`.

::mcp-tools

## What an agent can do

- push a **session**: rounds of moves, each with the post, the draft and the reason;
- read the **outcomes**: done, skipped, ratings, notes and rewrites;
- change your settings, only where you permitted it, per agent and per area. Every such change is logged and can be undone.

## What it cannot do

Post for you. A move always ends with you, on X.

## Limits

The plan sets the size of a round and a session, and how many agent calls you have a month. Past the allowance the agent gets a clear answer and nothing is written.