NERVsystems/osmmcp

Built in Go with the mcp-go library, it uses stdio transport and implements 24+ composable geospatial tools—including polyline encoding/decoding, OSRM routing integration, and emissions enrichment—designed as functional primitives where tool outputs chain seamlessly as inputs to enable emergent LLM workflows. Beyond standard geocoding, it provides specialized analyzers for neighborhood livability, EV charging along routes, commute comparison across transport modes, and school/parking discovery, with uniform interfaces and precise error handling for reliable MCP client integration.