This project acts as a Model Context Protocol (MCP) wrapper, transforming standard FHIR Terminology Services operations into structured tools that large language models and AI assistants can reliably invoke. It doesn't store code systems or ValueSets itself; instead, it intelligently forwards requests to an existing FHIR terminology backend such as `tx.fhir.org`, HAPI FHIR, or Ontoserver, and then normalizes the results for the client. This approach provides a familiar interface for AI applications, allowing them to resolve clinical codes without hard-coding HTTP clients, while reusing your current terminology infrastructure and maintaining a clear separation of concerns.
