01Seven dedicated MCP tools for data discovery, metadata exploration, and data retrieval.
02Two-layer caching (in-memory and persistent disk) to minimize API calls and enhance speed.
03Comprehensive metadata retrieval, combining structure, constraints, and codelist descriptions in a single call.
04Configurable dataflow blacklist to filter out specific or problematic datasets from all queries.
05Automated rate limiting and exponential backoff retry logic for robust API interaction.
061 GitHub stars