skills

Design VictoriaMetrics stream aggregation rules to reduce cardinality, sample frequency, or query load. Use whenever the user wants to apply, plan, or troubleshoot stream aggregation (`-streamAggr.config`, `-remoteWrite.streamAggr.config`), asks how to aggregate a high-cardinality metric, mentions downsampling at vmagent, replacing recording rules with stream aggregation, picking the right aggregation output (`total`, `rate_sum`, `sum_samples`, `quantiles`, `histogram_bucket`, etc.), choosing an aggregation interval, deciding where to place vmagent in the delivery pipeline, or whether to scale vmagent for aggregation. Also triggers on "aggregate histograms", "drop labels via vmagent", "reduce ingested samples", "pre-aggregate at vmagent", "stream aggregation accuracy", and "stream aggregation duplicates in cluster".

Use when investigating issues, debugging problems for applications, or responding to alerts in the Kubernetes cluster using VictoriaMetrics, VictoriaLogs, or VictoriaTraces.

Analyze VictoriaMetrics time series cardinality to find optimization opportunities — unused metrics, high-cardinality labels, problematic label values, histogram bloat. Produces actionable report with relabeling and stream aggregation recommendations. Use whenever the user mentions cardinality analysis, series reduction, unused metrics, high cardinality labels, TSDB optimization, storage cost reduction, metric cleanup, too many time series, or wants to reduce cardinality. Also trigger when discussing relabeling strategies, streaming aggregation opportunities, or "which metrics can we drop".

Find unused and rarely-queried metrics in VictoriaMetrics using the metric_names_stats API, then suggest optimization actions (drop rules, relabel configs). Use this skill when the user wants to find unused metrics, identify wasted storage, optimize metric ingestion, reduce cardinality by dropping unneeded metrics, clean up scrape targets, or asks about which metrics are never queried. Also trigger when the user mentions "metric cleanup", "unused series", "what metrics can I drop", "metric optimization", "wasted ingestion", or wants to reduce VictoriaMetrics resource consumption by eliminating unnecessary metrics.

Query AlertManager via curl using the v2 API. Use when listing active/silenced alerts, creating/managing silences, or checking alert inhibition state. Triggers on: alertmanager, silences, silence alerts, alert filters, alert inhibition, alertmanager API, create silence, delete silence.

Query VictoriaLogs via curl. Use when searching logs with LogsQL, running log stats queries, discovering log fields/streams, analyzing log hit patterns, or exploring log facets. Triggers on: log queries, LogsQL, log search, log stats, field discovery, stream discovery, log facets, log hits, log field values.

Query VictoriaMetrics metrics via curl. Use when running PromQL/MetricsQL queries, discovering metrics/labels, checking alerts and rules, inspecting TSDB status, exporting raw data, checking metric usage statistics, or debugging relabeling/downsampling/retention configs. Triggers on: metric queries, PromQL, MetricsQL, label discovery, series exploration, cardinality checks, alert status, recording rules, active/top queries, export data, metric statistics, relabel debug, downsampling debug, retention debug, flags.

Query VictoriaTraces via curl using the Jaeger-compatible API. Use when discovering traced services and operations, searching traces by service/operation/duration/tags, retrieving traces by ID, or mapping service dependencies. Triggers on: trace queries, span search, trace ID lookup, service discovery, operation discovery, service dependencies, distributed tracing, Jaeger API.

Analyze VictoriaMetrics query trace JSON to diagnose slow queries and produce a structured performance report with time breakdown, bottleneck analysis, and optimization recommendations. ALWAYS use this skill when: (1) the user mentions a VictoriaMetrics or VM trace, query trace, or trace JSON, (2) the user provides or references a JSON file containing duration_msec/message/children fields, (3) the user asks why a VictoriaMetrics/VM query is slow and has trace output, (4) the user asks about vmstorage node distribution, cache misses, or rollup performance in the context of a trace, (5) the user mentions vmselect trace, trace=1, or query performance debugging with VictoriaMetrics. This skill provides a structured report template that ensures consistent, thorough analysis — do not attempt to analyze VM traces without it.