Angi MCP Server — Live Demo

A working Model Context Protocol (MCP) server demonstrating how Angi integrates with LLM platforms.

🎯 Quick Start

🤖 ChatGPT Demo: https://chatgpt.com/g/g-699cf50686808191b07ade15b4b7284d-home-services-assistant-mcp-demo (ChatGPT Plus required)

🌐 Web Demo: https://piprads.github.io/angi-mcp/

💬 ChatGPT Setup: Guide | Demo Docs

📖 Full Documentation: DEMO.md — Complete technical guide

🚀 MCP Server: https://angi-mcp-production.up.railway.app

🛠️ Three MCP Tools

This demo implements three tools that address homeowner use cases:

1. search_professionals — Find verified Angi pros by category, zip code, and availability
2. get_home_advice — Answer home improvement questions with cost estimates and DIY guidance
3. request_quote — Submit quote requests to professionals (lead generation)

---

Deploy to Railway (10 minutes)

Step 1 -- Push to GitHub

cd angi-mcp
git init
git add .
git commit -m "Initial Angi MCP server"
# Create a new repo on github.com, then:
git remote add origin https://github.com/YOUR_USERNAME/angi-mcp.git
git push -u origin main

Step 2 -- Deploy on Railway

1. Go to railway.app and sign in with GitHub
2. Click New Project -> Deploy from GitHub repo
3. Select your angi-mcp repo
4. Railway auto-detects the config and deploys (~2 minutes)
5. Go to Settings -> Networking -> Generate Domain
6. Your public URL: https://your-app.railway.app

Step 3 -- Verify it's live

curl https://your-app.railway.app/health
# Should return: {"status":"ok","server":"angi-mcp","version":"1.0.0"}

---

Connect to Claude.ai (easiest)

1. Go to Claude.ai -> Settings -> Integrations
2. Click Add Integration
3. Enter: https://your-app.railway.app/mcp
4. Name it "Angi"
5. Done -- Claude can now call your tools in any conversation

Connect to Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "angi": {
      "url": "https://your-app.railway.app/mcp"
    }
  }
}

Restart Claude Desktop.

---

Tools Available

• search_professionals: Find verified Angi pros by category + zip code
• get_home_advice: Answer home improvement questions with cost estimates
• request_quote: Submit a quote request (the conversion/lead event)

---

Demo Scenarios

Scenario A -- Emergency plumbing: "My kitchen sink is clogged and backing up. I'm in zip 10002, I need someone today." LLM calls: get_home_advice -> search_professionals -> request_quote

Scenario B -- Remodel planning: "I want to remodel my bathroom. What should I expect to pay?" LLM calls: get_home_advice -> search_professionals

Scenario C -- AC not working: "My AC is blowing warm air. Can I fix this myself?" LLM calls: get_home_advice -> search_professionals

---

Project Structure

angi-mcp/
├── src/
│   ├── index.ts        # MCP server + 3 tools (StreamableHTTP transport)
│   └── mockData.ts     # Mock professionals + home advice Q&A
├── railway.toml        # Railway deploy config
├── package.json
├── tsconfig.json
└── README.md