MCP Server

Connect AI clients to Optionomics data

Optionomics runs a Model Context Protocol (MCP) server, letting compatible AI clients call a defined set of read-only market-data tools. You can ask for quotes, options chains, flow, gamma exposure, unusual activity, news and events, insider and politician trades, trade ideas with their track record, and earnings analyses—without giving the client write access to your Optionomics account.

• Endpoint: https://optionomics.ai/mcp (Streamable HTTP)
• Requirements: A Vega subscription on Optionomics
• Registry manifest: https://optionomics.ai/.well-known/mcp/server.json

What You’ll Need

1. Email — the email address on your Vega account
2. API key — created in the developer console at Developers → API keys

Every client configuration sends these as headers:

X-USER-EMAIL: [email protected]
X-USER-TOKEN: your-api-key

Clients that can only send a single Authorization header can authenticate with standard bearer syntax instead, encoding both values:

Authorization: Bearer base64("[email protected]:your-api-key")

Keep API keys out of committed project files. Prefer user-level client config or environment variables; if you use a workspace config file, add it to .gitignore.

The fastest path: the developer console’s MCP clients tab generates the exact config for your client with your email pre-filled, offers one-click install for clients that support it, and shows a Seen recently panel so you can confirm the connection end to end.

Available Tools

Once connected, your AI client can call these tools (this is the same list the reference page generates from the live server):

Available resources:

Every tool is read-only. Each call spends from the same 1,000 requests per minute account allowance as the REST API.

Claude Desktop

Edit your Claude Desktop config file (Claude Desktop → Settings → Developer → Edit Config):

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "optionomics": {
      "type": "http",
      "url": "https://optionomics.ai/mcp",
      "headers": {
        "X-USER-EMAIL": "[email protected]",
        "X-USER-TOKEN": "paste-your-api-key"
      }
    }
  }
}

Restart Claude Desktop completely (quit and reopen, not just close the window), then confirm the tools appear in the tools menu.

Claude Code

Create .mcp.json in your project root (checked in beside the code that uses it—but keep the key itself in an ignored file or environment variable):

{
  "mcpServers": {
    "optionomics": {
      "type": "http",
      "url": "https://optionomics.ai/mcp",
      "headers": {
        "X-USER-EMAIL": "[email protected]",
        "X-USER-TOKEN": "paste-your-api-key"
      }
    }
  }
}

Cursor

Use the Install in Cursor one-click button on the MCP clients tab, or create ~/.cursor/mcp.json (or .cursor/mcp.json for one project, excluded from source control):

{
  "mcpServers": {
    "optionomics": {
      "type": "http",
      "url": "https://optionomics.ai/mcp",
      "headers": {
        "X-USER-EMAIL": "[email protected]",
        "X-USER-TOKEN": "paste-your-api-key"
      }
    }
  }
}

Restart Cursor after saving.

VS Code (GitHub Copilot)

Use the Install in VS Code one-click button on the MCP clients tab, or create .vscode/mcp.json (VS Code uses a servers key):

{
  "servers": {
    "optionomics": {
      "type": "http",
      "url": "https://optionomics.ai/mcp",
      "headers": {
        "X-USER-EMAIL": "[email protected]",
        "X-USER-TOKEN": "paste-your-api-key"
      }
    }
  }
}

Then open the Command Palette (Cmd+Shift+P / Ctrl+Shift+P), run MCP: List Servers, and start the Optionomics server. Add .vscode/mcp.json to .gitignore if the file contains credentials.

Other clients

Any MCP client that speaks Streamable HTTP and can send either custom headers or a bearer token works the same way: point it at https://optionomics.ai/mcp with the X-USER-EMAIL/X-USER-TOKEN headers, or Authorization: Bearer base64(email:api-key) where only one header is possible. Client settings change independently of Optionomics, so follow your client’s current documentation for where its MCP config lives.

Registry-aware clients can also discover the server from its public manifest at https://optionomics.ai/.well-known/mcp/server.json.

Verify and debug your connection

• MCP console (/developers/console) — run any tool from the browser as your own account and read exactly the JSON a client would get back, plus the raw request to reproduce it.
• MCP logs (/developers/logs) — every call your clients made: tool, arguments, response, latency, and outcome.
• Seen recently (/developers/clients) — which client applications have actually called with your keys.

Example Prompts

Once connected, try these prompts in your AI client:

• “What’s the current put/call ratio and IV rank for SPY?”
• “Show me the top 10 unusual options activity alerts today, excluding ETFs”
• “Get the options chain for TSLA with calls only, sorted by strike”
• “What are the dark pool support and resistance levels for AAPL?”
• “Any congressional trades in NVDA this quarter?”
• “Summarize the latest earnings analysis for MSFT”
• “What’s the trade-idea track record for the last 90 days?”

Troubleshooting

Security Notes

• Never commit your API key to version control. Use environment variables or .gitignore‘d config files.
• Your API key authenticates as your user account. Treat it like a password, and revoke it from the developer console if it leaks.
• All connections are encrypted over HTTPS/TLS.
• Every call is logged to your own MCP log, so you can audit what a connected client actually asked for.