Skip to main content
Glama

MCP Starter for Puch AI

This is a starter template for creating your own Model Context Protocol (MCP) server that works with Puch AI. It comes with ready-to-use tools for job searching and image processing.

What is MCP?

MCP (Model Context Protocol) allows AI assistants like Puch to connect to external tools and data sources safely. Think of it like giving your AI extra superpowers without compromising security.

Related MCP server: Jobo Job Search MCP Server

What's Included in This Starter?

🎯 Job Finder Tool

  • Analyze job descriptions - Paste any job description and get smart insights

  • Fetch job postings from URLs - Give a job posting link and get the full details

  • Search for jobs - Use natural language to find relevant job opportunities

🖼️ Image Processing Tool

  • Convert images to black & white - Upload any image and get a monochrome version

🔐 Built-in Authentication

  • Bearer token authentication (required by Puch AI)

  • Validation tool that returns your phone number

Quick Setup Guide

Step 1: Install Dependencies

First, make sure you have Python 3.11 or higher installed. Then:

# Create virtual environment
uv venv

# Install all required packages
uv sync

# Activate the environment
source .venv/bin/activate

Step 2: Set Up Environment Variables

Create a .env file in the project root:

# Copy the example file
cp .env.example .env

Then edit .env and add your details:

AUTH_TOKEN=your_secret_token_here
MY_NUMBER=919876543210

Important Notes:

  • AUTH_TOKEN: This is your secret token for authentication. Keep it safe!

  • MY_NUMBER: Your WhatsApp number in format {country_code}{number} (e.g., 919876543210 for +91-9876543210)

Step 3: Run the Server

cd mcp-bearer-token
python mcp_starter.py

You'll see: 🚀 Starting MCP server on http://0.0.0.0:8086

Step 4: Make It Public (Required by Puch)

Since Puch needs to access your server over HTTPS, you need to expose your local server:

  1. Install ngrok: Download from https://ngrok.com/download

  2. Get your authtoken:

  3. Start the tunnel:

    ngrok http 8086

Option B: Deploy to Cloud

You can also deploy this to services like:

  • Railway

  • Render

  • Heroku

  • DigitalOcean App Platform

How to Connect with Puch AI

  1. Open Puch AI in your browser

  2. Start a new conversation

  3. Use the connect command:

    /mcp connect https://your-domain.ngrok.app/mcp your_secret_token_here

Debug Mode

To get more detailed error messages:

/mcp diagnostics-level debug

Customizing the Starter

Adding New Tools

  1. Create a new tool function:

    @mcp.tool(description="Your tool description")
    async def your_tool_name(
        parameter: Annotated[str, Field(description="Parameter description")]
    ) -> str:
        # Your tool logic here
        return "Tool result"
  2. Add required imports if needed

📚 Additional Documentation Resources

Official Puch AI MCP Documentation

  • Main Documentation: https://puch.ai/mcp

  • Protocol Compatibility: Core MCP specification with Bearer & OAuth support

  • Command Reference: Complete MCP command documentation

  • Server Requirements: Tool registration, validation, HTTPS requirements

Technical Specifications

Supported vs Unsupported Features

✓ Supported:

  • Core protocol messages

  • Tool definitions and calls

  • Authentication (Bearer & OAuth)

  • Error handling

✗ Not Supported:

  • Videos extension

  • Resources extension

  • Prompts extension

Getting Help


Happy coding! 🚀

Use the hashtag #BuildWithPuch in your posts about your MCP!

This starter makes it super easy to create your own MCP server for Puch AI. Just follow the setup steps and you'll be ready to extend Puch with your custom tools!

Related MCP Connectors

Related MCP Servers

  • A
    license
    Not graded
    quality
    C
    maintenance
    Enables AI agents to search and explore remote job listings, salary stats, and popular roles, and supports token-authenticated actions like job alerts and apply redirects.
    MIT
  • A
    license
    Not graded
    quality
    C
    maintenance
    Enables LLM clients to search Jobo's live job index and fetch detailed job postings, with canonical search and fetch tools for ChatGPT Deep Research compatibility. Supports OAuth 2.1 authentication, structured filters, and self-hosting.
    8 npm
    MIT
  • A
    license
    Not graded
    quality
    C
    maintenance
    Exposes the nofluffjobs.com public job board as two standard MCP tools, letting any MCP host search listings by keyword, location, posting date, and remote preference, then fetch full details for a chosen posting. It is read-only, requires no Zhencheng runtime, and returns normalized fields such as title, company, apply URL, and snippet.
    MIT
  • A
    license
    Not graded
    quality
    C
    maintenance
    Enables AI agents to search a public job board by keyword, location, remote preference, and posting date, then fetch full details and apply links for individual listings. It wraps the board's public API/RSS as read-only stdio tools usable from any MCP host.
    MIT