halans/evlife-mcp-server
If you are the rightful owner of evlife-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.
The EV Life Content MCP Server provides content querying tools for electricvehicle.life, offering both local and cloud deployment options.
EV Life Content MCP Server
This MCP (Model Context Protocol) server provides content querying tools for electricvehicle.life. It offers two deployment options: a local stdio-based server for direct Claude Desktop integration, and a Cloudflare Workers deployment for remote access.
Available Tools
The MCP server provides four content querying tools:
- search_content - Search for specific terms in electricvehicle.life content with context
- get_section - Retrieve specific sections by title/heading
- get_full_content - Get the complete content with optional truncation
- get_content_summary - Generate content statistics and table of contents
Quick Start
Option 1: Local Stdio Server (Recommended)
- Clone and install dependencies:
git clone <your-repo-url>
cd evlife-mcp-server
npm install
- Configure Claude Desktop by adding to your MCP config:
{
"mcpServers": {
"evlife-content": {
"command": "node",
"args": ["/path/to/evlife-mcp-server/mcp-stdio.js"],
"env": {}
}
}
}
- Restart Claude Desktop and the tools will be available.
Option 2: Cloudflare Workers Deployment
-
Prerequisites:
- Sign up for a Cloudflare account
- Install Wrangler CLI
-
Authenticate with Cloudflare:
npx wrangler login
- Deploy to Cloudflare Workers:
npm run deploy
This will deploy your MCP server to a URL like: evlife-mcp-server.<your-account>.workers.dev/sse
Cloudflare Workers Features
This MCP server leverages several Cloudflare Workers features:
- Edge Computing: Runs close to users worldwide for low latency
- Durable Objects: Maintains stateful MCP agent instances
- Server-Sent Events: Real-time communication with MCP clients
- No Cold Starts: Fast response times with Cloudflare's V8 isolates
- Built-in Observability: Monitoring and analytics through Cloudflare dashboard
Worker Configuration
The server exposes two endpoints:
/sse- Server-Sent Events endpoint for MCP communication/mcp- Standard MCP endpoint
Configuration is managed in wrangler.jsonc:
{
"name": "evlife-mcp-server",
"main": "src/index.ts",
"compatibility_date": "2025-03-10",
"durable_objects": {
"bindings": [
{
"class_name": "MyMCP",
"name": "MCP_OBJECT"
}
]
}
}
Development
Local Development with Wrangler
# Start the Cloudflare Workers dev server
npm run dev
# Run TypeScript type checking
npm run type-check
# Format code
npm run format
# Fix linting issues
npm run lint:fix
Local Development with Stdio Server
# Test the stdio server directly
node mcp-stdio.js
# The server will wait for MCP protocol messages on stdin
Project Structure
├── src/
│ └── index.ts # Cloudflare Workers MCP server
├── mcp-stdio.js # Local stdio MCP server
├── package.json # Dependencies and scripts
├── wrangler.jsonc # Cloudflare Workers config
├── CLAUDE.md # Claude-specific documentation
└── README.md # This file
Content Source
The server fetches content from https://electricvehicle.life/llms-full.txt and caches it for 5 minutes to improve performance. The content includes:
- Personal blog posts and essays
- Technical documentation
- Project descriptions
- Conference notes and insights
Test locally with MCP Inspector
npx @modelcontextprotocol/inspector
Connecting to Claude Desktop
For Local Stdio Server
Add this to your Claude Desktop MCP configuration:
{
"mcpServers": {
"evlife-content": {
"command": "node",
"args": ["/absolute/path/to/mcp-stdio.js"],
"env": {}
}
}
}
For Deployed Workers (with mcp-remote)
{
"mcpServers": {
"evlife-content": {
"command": "npx",
"args": [
"mcp-remote",
"https://your-worker-url.workers.dev/sse"
]
}
}
}
Connect to Cloudflare AI Playground
For deployed Workers, you can test the MCP server using Cloudflare AI Playground:
- Go to https://playground.ai.cloudflare.com/
- Enter your deployed MCP server URL (
your-worker-url.workers.dev/sse) - Test the content querying tools directly
Deployment Customization
Custom Domain (Optional)
To use a custom domain with your Cloudflare Worker:
- Add a custom domain in your Cloudflare dashboard
- Update
wrangler.jsoncwith your domain:
{
"routes": [
{
"pattern": "mcp.yourdomain.com/*",
"custom_domain": true
}
]
}
Environment Variables
Add environment variables for configuration:
# Set environment variables
npx wrangler secret put API_KEY
npx wrangler secret put CONTENT_URL
Then access them in your Worker:
// In src/index.ts
const contentUrl = env.CONTENT_URL || "https://electricvehicle.life/llms-full.txt";
Troubleshooting
Common Issues
-
MCP Connection Failed
- Ensure the server URL is correct
- Check that the Worker is deployed and accessible
- Verify Claude Desktop configuration
-
Content Fetching Errors
- Check if
https://electricvehicle.life/llms-full.txtis accessible - Verify network connectivity from the Worker
- Monitor Cloudflare logs for fetch errors
- Check if
-
Development Server Issues
- Run
npm installto ensure dependencies are installed - Check Node.js version compatibility (v18+ recommended)
- Use
npm run type-checkto identify TypeScript errors
- Run
Monitoring
- View Worker logs in the Cloudflare dashboard
- Use
wrangler tailfor real-time log monitoring - Monitor MCP server logs in Claude Desktop
Customization
To add your own tools:
- Stdio server: Edit
mcp-stdio.jsand add new tool handlers - Workers server: Edit
src/index.tsand add tools in theinit()method
Example: Adding a New Tool
// In src/index.ts or mcp-stdio.js
this.server.tool(
"new_tool",
{
parameter: z.string().describe("Tool parameter")
},
async ({ parameter }) => {
// Tool implementation
return {
content: [{ type: "text", text: `Result: ${parameter}` }]
};
}
);