# Connect your AI assistant to Luca

Claude, ChatGPT, Cursor, VS Code, or any MCP client can work your review queue, look up leads, and draft replies with you. Luca is the product; SetLuca is the company that builds it.

> New connections start with read + draft access only. Nothing sends without your review.

## Connectors: Claude and ChatGPT

Both assistants take Luca as a connector. Open the connector settings, add a custom connector with this URL, then sign in to Luca and choose what the assistant may do on the consent screen:

`https://mcp.setluca.com/mcp`

- [Add to Claude](https://claude.ai/settings/connectors), then choose Add custom connector.
- [Add to ChatGPT](https://chatgpt.com/#settings/Connectors), then choose Create under Connectors.

## MCP server, hosted: Cursor, VS Code, and any remote MCP client

The connectors sit on Luca's MCP server, and any client that speaks remote MCP over Streamable HTTP can use the same URL directly. One-click install:

- [Add to Cursor](cursor://anysphere.cursor-deeplink/mcp/install?name=luca&config=eyJ1cmwiOiJodHRwczovL21jcC5zZXRsdWNhLmNvbS9tY3AifQ==)
- [Add to VS Code](vscode:mcp/install?%7B%22name%22%3A%22luca%22%2C%22url%22%3A%22https%3A%2F%2Fmcp.setluca.com%2Fmcp%22%7D)

## MCP server, local: Claude Desktop, Claude Code, and any stdio client (API key)

1. Create an API key in **Luca -> Settings -> Developer API keys**.
2. Add this to your MCP client configuration:

```
{
  "mcpServers": {
    "luca": {
      "command": "npx",
      "args": [
        "-y",
        "@setluca/mcp"
      ],
      "env": {
        "LUCA_API_KEY": "<your-key>"
      }
    }
  }
}
```

Or with the Claude Code CLI:

```
claude mcp add luca --env LUCA_API_KEY=<your-key> -- npx -y @setluca/mcp
```

Then ask your assistant "What's in my Luca review queue?" to confirm it works. Never paste your API key into a chat. It belongs in the config file only.

Package: `@setluca/mcp@0.2.1` - Registry: `io.github.setluca/luca-mcp` - Developers: see [/developers](/developers).