Livestorm API MCP Server

This project creates a Model Context Protocol (MCP) server that wraps the Livestorm API, exposing:

• GET endpoints as Resources
• POST, PUT, DELETE endpoints as Tools

1. The server fetches and parses the Livestorm API's OpenAPI definition
2. It dynamically creates MCP Resources and Tools based on the API endpoints
3. When a client requests a Resource or Tool, the server proxies the request to the Livestorm API

When defining the MCP server in your LLM client applications use the following snippet:

{
  "mcpServers": {
    "livestorm-api": {
      "type": "streamable-http",
      "url": "https://supagroova--7fab7ae4322911f080e9569c3dd06744.web.val.run/mcp",
      "note": "For Streamable HTTP connections, add this URL directly in Client"
    }
  }
}

{
  "mcpServers": {
    "livestorm-api": {
      "type": "sse",
      "serverUrl": "https://supagroova--7fab7ae4322911f080e9569c3dd06744.web.val.run/sse"
    }
  }
}

• index.ts: Main entry point with HTTP trigger
• livestorm.ts: Functions to fetch and parse the OpenAPI definition
• mcp.ts: MCP server setup and configuration

You can store your environment variables in a .env file in the project root. These will be loaded automatically when you run the server (using Deno std/dotenv).

Example .env:

RUN_LOCAL=1
PORT=8787
LIVESTORM_API_TOKEN=your-livestorm-api-token-here

This is useful for API tokens and local configuration.

You can run this MCP server locally using the Deno runtime, which supports TypeScript and npm:/URL imports out of the box.

With Homebrew (macOS):

brew install deno

In your project directory, run:

deno task dev

To specify a custom port:

PORT=3000 deno task dev

Deno will automatically fetch and cache all dependencies.