Orchestrates live flight data from AeroAPI and BigQuery catalog information using local LLM inference via LangGraph.
Sponsored
Aero Ref serves as a local agent and Model Context Protocol (MCP) demonstration, seamlessly integrating real-time flight information from FlightAware AeroAPI with a Google BigQuery catalog of demo airports. It leverages LangGraph for orchestrating complex agentic workflows and utilizes Ollama for local Large Language Model inference. The system runs as a multi-server MCP agent, demonstrating how different services can expose tools and prompts, enabling dynamic, intelligent interactions for flight data queries and comparisons.
주요 기능
011 GitHub stars
02MCP prompts for server-side workflow templates and guardrails
03Local LLM inference via Ollama (e.g., Qwen2.5) and LangChain
04Multi-server MCP agent architecture (Flight and BigQuery servers)
05Integration with FlightAware AeroAPI for live flight data
06Google BigQuery integration for airport catalog information
07Orchestration of agentic workflows using LangGraph
사용 사례
01Demonstrating multi-server agentic AI interactions with external APIs
02Developing and testing applications utilizing the Model Context Protocol (MCP)
03Building flight intelligence systems with real-time and catalog data
04Exploring LangGraph orchestration with local LLMs and custom tools
