Demo Local LLM

About

This demo project illustrates the creation of MCP clients and servers using both Python (with FastAPI) and TypeScript, showcasing integration with a local Large Language Model (LLM). It outlines a workflow where the client requests tools from the server, the LLM intelligently selects a tool based on a given prompt, and the client then executes the chosen tool on the server. The project highlights its versatility by supporting all combinations of Python and TypeScript clients and servers for a comprehensive demonstration.

Key Features

• Multi-language MCP client implementations (Python & TypeScript).
• Local LLM integration via Ollama for tool selection.
• Multi-language MCP server implementations (Python/FastAPI & TypeScript).
• Demonstrates LLM-driven tool selection and execution workflow.
• Ensures LLM returns correctly typed arguments for tool calls.
• 0 GitHub stars

Use Cases

• Showcasing cross-language communication in microservices-like environments.
• Exploring local LLM integration for function calling and tool execution.
• Learning and prototyping MCP client-server architectures.