JetBrains AI Assistant

Remote MCP support with no header field — so a keyed endpoint like this one goes through the bridge.

Setup

1. Create a key in the console and export it as UNZOI_KEY.
2. Open Settings → Tools → AI Assistant → Model Context Protocol (MCP) and add a new server.
3. Paste the config below — this is the bridge form, not a direct url entry.
4. Ask the assistant something that needs current reporting and check the tool call in its run log.

{
  "mcpServers": {
    "unzoi": {
      "command": "npx",
      "args": [
        "-y", "mcp-remote", "https://api.unzoi.com/mcp",
        "--header", "x-api-key:${UNZOI_KEY}"
      ],
      "env": { "UNZOI_KEY": "nai_your_key_here" }
    }
  }
}

What differs here

• AI Assistant's remote MCP entry takes only a url — there is no documented field for a custom header, so a header-authenticated endpoint like this one cannot be added the direct way an unauthenticated or OAuth-protected server can.
• The mcp-remote bridge above is a command-based entry, so it goes through the same MCP panel as any local server — JetBrains does not distinguish bridged and native servers in the UI.
• IntelliJ IDEA, PyCharm, WebStorm, Rider and the rest of the JetBrains lineup share this panel, so the same configuration works across the family, not only in IntelliJ IDEA.

Next

• Get a key at console.unzoi.com, if you have not.
• The tool reference — what the model is reading when it chooses between the 18 tools.
• Stories for context windows — the one habit that most changes how well an agent works against this API.
• Limits — an agent that loops can spend a month's credits faster than you expect.
• JetBrains AI Assistant's own MCP documentation — authoritative for anything about the client itself, including config paths, which move.

Endpoint, for copying: https://api.unzoi.com/mcp