elasticsearch-esql

name	elasticsearch-esql
description	Use this skill when writing or debugging ES|QL queries for Elasticsearch. Activate when the user asks to query logs, metrics, traces, or any Elasticsearch data using ES|QL syntax.
metadata	{"version":"0.1.0","visibility":"public"}

Elasticsearch ES|QL Query Authoring

ES|QL Basics

ES|QL (Elasticsearch Query Language) is a piped query language for filtering, transforming, and aggregating Elasticsearch data.

Syntax

FROM <index-pattern>
| WHERE <condition>
| STATS <aggregation> BY <field>
| SORT <field> [ASC|DESC]
| LIMIT <n>

Running Queries

Use the elastic CLI:

elastic es query 'FROM logs-* | WHERE @timestamp > NOW() - 1 HOUR | LIMIT 10'

Common Patterns

Filter by time range:

FROM logs-*
| WHERE @timestamp > NOW() - 24 HOURS

Count by field:

FROM logs-*
| STATS count = COUNT(*) BY service.name
| SORT count DESC

Percentiles:

FROM metrics-apm*
| STATS p50 = PERCENTILE(transaction.duration.us, 50),
        p95 = PERCENTILE(transaction.duration.us, 95),
        p99 = PERCENTILE(transaction.duration.us, 99)
  BY service.name

Time bucketing:

FROM logs-*
| WHERE log.level == "error"
| STATS errors = COUNT(*) BY bucket = BUCKET(@timestamp, 5 minute)
| SORT bucket

Multi-field filtering:

FROM logs-*
| WHERE service.name == "api-gateway" AND http.response.status_code >= 500
| KEEP @timestamp, message, http.response.status_code, trace.id
| SORT @timestamp DESC
| LIMIT 50

Type Functions

• TO_STRING(field), TO_INTEGER(field), TO_DOUBLE(field) -- type conversions
• DATE_TRUNC(interval, field) -- truncate timestamps
• CONCAT(a, b) -- string concatenation
• LENGTH(field) -- string length
• TRIM(field), LEFT(field, n), RIGHT(field, n) -- string manipulation

Aggregation Functions

• COUNT(*), COUNT(field), COUNT_DISTINCT(field)
• SUM(field), AVG(field), MIN(field), MAX(field)
• PERCENTILE(field, pct), MEDIAN(field)
• VALUES(field) -- collect distinct values

Tips

• Always specify a time range with WHERE @timestamp > NOW() - <duration> to avoid scanning too much data
• Use LIMIT to control output size
• Use KEEP to select specific columns before output
• Use DROP to exclude columns
• Use RENAME old AS new to rename columns
• Use EVAL new_field = expression to create computed columns
• Use DISSECT or GROK for parsing unstructured text fields