Skip to main content

Cornix MCP: Connect Cornix to Your AI Assistant

Cornix MCP is currently in beta. Learn how to connect Cornix to an MCP-compatible AI assistant and manage your portfolio, trades, and bots using natural language.

A
Written by Adam Zidiker

Cornix MCP connects your Cornix account to compatible AI assistants, allowing you to explore your portfolio and manage trades and bots using natural language.

💡 Note: Cornix MCP is currently in beta. If you do not see MCP in your Cornix dashboard, the feature may not yet be enabled for your account.

Go to Your Cornix MCP Page

What Is Cornix MCP?

Model Context Protocol (MCP) is an open standard that lets compatible AI assistants securely connect to external services such as Cornix. Once connected, your assistant can use the tools made available by Cornix while respecting your existing Cornix permissions.

What You Can Do with Cornix MCP

See Your Portfolio

View your connected accounts and get a clear summary of your current portfolio.

Example: “Summarize my portfolio.”

Manage Trades

Review open and closed trades, inspect trade details, and create, edit, or cancel trades.

Example: “Show my open BTC trades.”

Control Your Bots

Review, create, edit, or delete bots, and start or pause them whenever you need to.

Example: “Pause my DCA bot.”

How to Connect Cornix to Your AI Assistant

Follow these steps to connect your preferred AI client:

  1. Sign in to the Cornix dashboard and open AI → MCP.

  2. Copy the URL shown under Your Cornix MCP server.

  3. Choose your AI client below and add Cornix as a remote HTTP MCP server.

  4. Complete the Cornix sign-in and approval flow in your browser.

Client-Specific Setup Instructions

Claude Code

  1. Open your terminal.

  2. Run the following command, using the production Cornix MCP URL:

    claude mcp add --transport http cornix https://dashboard.cornix.io/app/mcp

  3. Open Claude Code and enter /mcp.

  4. Select Cornix and complete the sign-in flow in your browser.

Cursor

  1. Open or create ~/.cursor/mcp.json.

  2. Add the following configuration, using the production Cornix MCP URL:

    {
      "mcpServers": {
        "cornix": {
          "url": "https://dashboard.cornix.io/app/mcp"
        }
      }
    }

  3. Save the file and restart Cursor.

  4. Select Cornix in your MCP settings and complete the sign-in flow.

VS Code

  1. Open the Command Palette and run MCP: Open User Configuration.

  2. Add the following server to mcp.json, using the production Cornix MCP URL:

    {
      "servers": {
        "cornix": {
          "type": "http",
          "url": "https://dashboard.cornix.io/app/mcp"
        }
      }
    }

  3. Save the file, run MCP: List Servers, and start Cornix.

  4. Approve the server when prompted and complete the sign-in flow.

Codex

  1. Open your terminal.

  2. Run the following commands, using the production Cornix MCP URL:

    codex mcp add cornix --url https://dashboard.cornix.io/app/mcp
    codex mcp login cornix

  3. Complete the Cornix sign-in flow in the browser window that opens.

  4. Restart any running Codex session. Cornix will then be available in the CLI, IDE extension, and desktop app on this host.

Gemini CLI

  1. Open or create ~/.gemini/settings.json.

  2. Add the following configuration, using the production Cornix MCP URL:

    {
      "mcpServers": {
        "cornix": {
          "httpUrl": "https://dashboard.cornix.io/app/mcp"
        }
      }
    }

  3. Save the file and restart Gemini CLI.

  4. Enter /mcp auth cornix and complete the sign-in flow in your browser.

Security and Permissions

Your Cornix password is never shared with your AI client. Cornix uses secure authorization, and the assistant can only access actions and information allowed by your Cornix permissions.

Because Cornix MCP can perform account-changing actions, carefully review the assistant’s request and the intended parameters before approving trades, bot changes, or other actions.

AI Integration Disclaimer

Cornix provides the connection and tools, but your chosen AI service generates and is responsible for its responses. AI output may be inaccurate or incomplete. Always review every response before approving trades, bot changes, or other actions.

Did this answer your question?