MCP Starter

Starter code to add MCPs to Puch AI #BuildWithPuch

typescript

60

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "turboml-inc-mcp-starter": {
      "command": "python",
      "args": [
        "mcp_starter.py"
      ],
      "env": {
        "MY_NUMBER": "919876543210",
        "AUTH_TOKEN": "your_secret_token_here"
      }
    }
  }
}

You can extend Puch AI with a dedicated MCP server that safely connects your AI assistant to external tools and data sources. This starter gives you a ready-to-run Python MCP server with bearer-token authentication and simple tooling to process text and images, plus a sample user-scoped task server to illustrate multi-user usage.

How to use

You will run a local MCP server and connect Puch AI to it so your assistant can call tools you expose. Start the server, expose it publicly if required by your setup, and then instruct Puch AI to connect to the endpoint using the secret token you configured.

How to install

Prerequisites you need on your machine before starting:

Python 3.11 or higher.

Step-by-step setup:

# Create a virtual environment
uv venv

# Install all required packages
uv sync

# Activate the environment
source .venv/bin/activate

Additional content

Configuration and run flow described here include using a bearer token for authentication and setting a WhatsApp-like number in the environment. The starter focuses on a bearer-token MCP server and provides example configurations to help you get started fast.

Configuration and security notes

AUTH_TOKEN is your secret token used for authentication. Keep it safe and do not share it.
MY_NUMBER should be your WhatsApp number in the form {country_code}{number} (for example 919876543210 for +91-9876543210).

Troubleshooting and diagnostics

If you need more detailed error information during development, enable the debug diagnostics mode in your MCP client and server setup to surface stack traces and helpful hints.

Available tools

echo_tool

Text input/output tool that echoes or processes text in a simple, echo-style manner to demonstrate text tool usage within MCP.

image_tool

Image input/output tool that processes images, for example converting to black and white or applying basic transformations.

task_manager

Sample MCP server demonstrating how to use a user-provided identifier (puch_user_id) to scope tasks and data per user in a multi-user context.