Agentic AI Starter Kit

This repository serves as a kit for learning how to build Agentic AI applications. Currently, it has code to demonstrate two key communication patterns: Agent-to-Agent (A2A) communication and Model Context Protocol (MCP) communication. We will add more concepts in the future.

Project Structure

• a2a/: Contains a hands-on demonstration of the Agent-to-Agent (A2A) communication pattern. This includes:

  • Detailed README.md for setup and running A2A demos.

• mcp/: Contains a demonstration of the Model Context Protocol (MCP) communication pattern. This includes:

  • clients/: Client demo application (mcp_client_demo.py).
  • servers/: Example MCP servers (e.g., math_server.py).
  • Detailed README.md for setup and running MCP demos.

• .env: Environment variables for configuring LLM access and other settings.

Prerequisites

• Python 3.12.7 or higher
• Basic familiarity with Python and command-line operations.

Setup Instructions

1. Clone the Repository

   git clone https://github.com/NihalJain/agentic-ai-learning-kit.git
   cd agentic-ai-learning-kit
   

2. Set Up Virtual Environment

   python3 -m venv .venv
   # On Windows:
   .venv\Scripts\Activate.ps1
   # On macOS/Linux:
   source .venv/bin/activate
   

3. Configure Environment Variables Create a .env file in the root directory of the project (agentic-ai--learning-kit) and populate it with the necessary LLM API keys and other configurations.

• LLM_MODEL_PROVIDER: (Required) Specifies the model provider (google or openai). Defaults to google, if not explicitly set.
• LLM_API_KEY: (Required) The API key for authenticating with the LLM provider.
• LLM_BASE_URL: (Optional) The base URL of the LLM API. Defaults to https://generativelanguage.googleapis.com/v1beta for google source, or https://api.openai.com/v1/ for openai source.
• LLM_MODEL_NAME: (Optional) The name of the LLM model to use. Defaults to gemini-2.5-pro for google source, or gpt-4o-mini for openai source.
• LLM_TEMPERATURE: (Optional) The temperature setting for the LLM. Defaults to 0.

Refer to the a2a/README.md and mcp/README.md for detailed information on required environment variables for each respective demo.

Running the Demos

To run the demos, navigate into the respective subdirectories and follow their README.md instructions:

• A2A Demos: See a2a/README.md for detailed instructions.
• MCP Demos: See mcp/README.md for detailed instructions.

Author Information

For any issues or queries, please contact the author:

• Name: Nihal Jain
• Email: nihaljain.cs@gmail.com