Agents MCP Usage
Createdandrewginns
Demonstrates the use of a Model Context Protocol (MCP) server with various agent frameworks like Google ADK, LangGraph, OpenAI Agents, and Pydantic-AI.
About
This repository serves as a practical demonstration of integrating a simple Model Context Protocol (MCP) server with several popular agent frameworks. It showcases how to leverage MCP for standardized interaction between agents built with Google ADK, LangGraph, OpenAI Agents, and Pydantic-AI, and various LLM providers. The included MCP server, enhanced with a datetime tool, enables seamless context sharing and tool usage across these frameworks, while Logfire integration provides comprehensive tracing and observability for LLM agent behavior.
Key Features
- Facilitates switching between different LLM providers without major code changes.
- Provides Logfire integration for tracing and monitoring LLM agent interactions.
- Offers example MCP tools like adding numbers and getting the current time.
- Demonstrates MCP integration with Google ADK, LangGraph, OpenAI Agents, and Pydantic-AI.
- 1 GitHub stars
- Includes a simple, locally runnable MCP server.
Use Cases
- Experimenting with different agent frameworks using a common MCP server.
- Implementing a consistent interface for tools and data in LLM applications.
- Standardizing agent interactions with LLMs across different frameworks.
