Elasticsearch MCP Server
Connects AI agents to Elasticsearch clusters via MCP, exposing tools to list indices, inspect mappings, execute Query DSL searches, run ES|QL queries, and inspect shard distributions. Note: officially deprecated in favor of Elastic Agent Builder's native MCP endpoint in Elastic 9.2+.
Score Breakdown
⚙ Agent Friendliness
🔒 Security
Elasticsearch MCP. API key with index-level permissions. Search can expose sensitive indexed data — scope carefully. TLS required in production. X-Pack security features.
⚡ Reliability
Best When
You are on Elastic 8.x and need a quick MCP bridge to query Elasticsearch from Claude Desktop or another MCP client without standing up additional infrastructure.
Avoid When
You are on Elastic 9.2+ or starting a new project; the project is officially deprecated and the native endpoint is preferred.
Use Cases
- • Natural language querying of Elasticsearch indices from an LLM client
- • Inspecting index structure and shard health during incident response
- • Running ES|QL analytics queries via conversational AI
Not For
- • Elasticsearch clusters below version 8.x
- • Write operations (indexing, creating/deleting indices)
- • New projects targeting Elastic 9.2+ (use the native Agent Builder MCP endpoint instead)
Interface
Authentication
Supports ES_API_KEY or ES_USERNAME/ES_PASSWORD via environment variables. SSL verification can be skipped with ES_SSL_SKIP_VERIFY=true (not recommended in production).
Pricing
Apache-2.0 open source. Requires a running Elasticsearch cluster (self-hosted or Elastic Cloud, which has its own costs).
Agent Metadata
Known Gotchas
- ⚠ Project is officially deprecated; new work should target Elastic 9.2+ native MCP endpoint.
- ⚠ Only read operations exposed; agents cannot index or modify data.
- ⚠ SSL verification bypass flag (ES_SSL_SKIP_VERIFY) is a security risk if used in production.
- ⚠ Streamable-HTTP/SSE modes require additional mcp-proxy installation for stdio bridging.
Alternatives
Full Evaluation Report
Detailed scoring breakdown, competitive positioning, security analysis, and improvement recommendations for Elasticsearch MCP Server.
Scores are editorial opinions as of 2026-03-06.