01Automatic rate limiting and exponential backoff retry logic for reliable API interaction
02Optimized workflow for quickly discovering datasets and retrieving complete metadata with descriptions
03Configurable dataflow blacklist to filter out specific or problematic datasets
04Seven dedicated MCP Tools for data discovery, metadata retrieval, and data fetching
05Two-layer caching (in-memory and persistent disk) to enhance performance and reduce API calls
060 GitHub stars