Skip to content

Elasticsearch MCP — Full-Text Search & Observability for AI Agents

What It Is

Elastic MCP exposes Elasticsearch search and observability capabilities over the Model Context Protocol. AI agents can query any index with Elasticsearch Query DSL, run kNN semantic search, and monitor APM traces, logs, and infrastructure metrics.

Tools Available

Tool Description
Full-text search Query any index with Elasticsearch Query DSL
Vector search kNN semantic search across embeddings
Aggregations Buckets, metrics, pipeline aggregations
Observability APM traces, logs, and infrastructure metrics

Quick Start

npx -y @elastic/mcp-server-elasticsearch

Business Use Cases

  1. Log investigation: "Find all ERROR-level logs from the payments service in the last hour"
  2. Semantic search: "Find documents related to our refund policy"
  3. Infrastructure health: "Show me APM traces with latency over 500ms"

Limitations

  • Elasticsearch 8.x+ required. Vector search needs dense_vector field mappings.

See Also