Elasticsearch Observability Skill

You have access to the Elasticsearch MCP server tools for querying and analyzing data from Elasticsearch clusters.

Available Tools

Workflow

1. Understand the request — Determine what Elasticsearch data the user needs
2. Discover indices — Use list_indices to find relevant data sources
3. Check mappings — Use get_mappings to understand available fields
4. Query data — Use search (Query DSL) or esql (ES|QL) to retrieve results
5. Check health — Use get_shards for cluster and index health information
6. Analyze results — Summarize findings with actionable recommendations

Best Practices

• Always start by listing indices to discover available data sources
• Check field mappings before writing queries to ensure correct field names
• Use ES|QL for complex aggregations and data exploration
• Use Query DSL search for precise filtering and full-text search
• Start with smaller timeframes and add LIMIT to optimize query performance
• Prefer safe, bounded queries — include explicit time filters
• Do not fabricate index names or field names; discover them first