You'd reach for this when Claude needs to search across large document collections or structured datasets beyond simple keyword matching. It exposes distributed search capabilities including full-text search, semantic queries that understand meaning rather than just exact matches, and structured data operations. Think of it as bringing Elasticsearch-style functionality into your MCP workflow. The distributed aspect means it can handle indexing and queries across multiple nodes, so it scales beyond what you'd get from grep or basic SQLite lookups. Useful when you're building retrieval systems, documentation search, or any project where Claude needs to dig through substantial data stores intelligently.