Agentic Travel Planner

About

This tool provides a production-ready, framework-free agentic workflow designed for efficient travel planning. Built from scratch with Python and the Model Context Protocol (MCP), it features a custom lightweight MCP Client/Server and supports multiple LLM providers, including OpenAI, Anthropic, and Google Gemini. Users can interact via an interactive CLI or a modern Flask-based web UI, benefiting from integrated tools for flight search, car rental, weather forecasts, and simulated payments. The project emphasizes production readiness with structured logging, Pydantic validation, robust error handling, state management, performance caching, comprehensive testing, and Docker support, also serving as an educational resource with an fully annotated codebase.

Key Features

• Custom MCP Client/Server for standardized tool communication with multi-LLM support.
• Framework-free agentic architecture built with standard Python.
• Production-ready features including structured logging, Pydantic validation, error handling with retries, and Docker support.
• Integrated tools for flight search, car rental, weather forecasts, and payment simulation.
• 0 GitHub stars
• Interactive CLI and a modern Flask-based Web UI with search history and real-time status updates.

Use Cases

• Streamlining personal and business travel arrangements using an AI agent.
• Providing a robust reference architecture and demonstration for agentic workflow development.
• Serving as an educational resource for understanding the Model Context Protocol (MCP) in practice through an annotated codebase.