agentleFS
SawanaLabsSawanaLabs/agent-demos/.agents/skills/victoriametrics-query/SKILL.md

victoriametrics-query

Skill4 starsChanged 57 days agoMITView on GitHub

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.

1 other project uses a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    All curl commands load auth from a curl config file - works for both remote and local:

    line 29
  2. Very specific:

    When VM_CURL_CONFIG is unset, curl reads /dev/null and sends no auth header.

    line 35
  3. Very specific:

    All times accept RFC3339 (2026-03-07T09:00:00Z) or Unix seconds (1709769600).

    line 238

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

VictoriaMetrics Metrics Query

Query VictoriaMetrics HTTP API directly via curl. Covers instant/range queries, label/series discovery, alerts, rules, instance diagnostics, raw data export, metric usage statistics, and config debugging tools.

Environment

# $VM_METRICS_URL - base URL
#   cluster: export VM_METRICS_URL="https://vmselect.example.com/select/0/prometheus"
#   single: export VM_METRICS_URL="http://localhost:8428"
# $VM_CURL_CONFIG - curl config file with auth header (set for remote, unset for local)
#   Remote: export VM_CURL_CONFIG="$HOME/.config/victoriametrics/curl.conf"
#   Local:  leave unset (defaults to /dev/null - no auth)

Auth Pattern

All curl commands load auth from a curl config file - works for both remote and local:

curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s "$VM_METRICS_URL/api/v1/query?query=up" | jq .

When VM_CURL_CONFIG is unset, curl reads /dev/null and sends no auth header. When set, it must point to a mode-0600 curl config file containing header = "Authorization: Bearer <token>". Never print its contents.

Core Endpoints

Instant Query

# Query at current time
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/query?query=up" | jq .

# Query at specific time
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/query?query=up&time=2026-03-07T09:00:00Z" | jq .

Parameters: query (required), time (optional, RFC3339 or Unix seconds), step, timeout

Range Query

curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/query_range?query=rate(http_requests_total[5m])&start=2026-03-07T00:00:00Z&end=2026-03-07T12:00:00Z&step=5m" | jq .

Parameters: query (required), start (required), end (optional), step (required). Times in RFC3339 or Unix seconds.

Labels Discovery

# All label names
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/labels" | jq '.data[]'

# Label values (label_name is a PATH parameter)
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/label/namespace/values" | jq '.data[]'

# Label values filtered by series matcher
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  --data-urlencode 'match[]={job="kubelet"}' \
  "$VM_METRICS_URL/api/v1/label/namespace/values" | jq '.data[]'

Series Discovery

# Find series matching selector
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  --data-urlencode 'match[]={namespace="myapp"}' \
  "$VM_METRICS_URL/api/v1/series?limit=20" | jq '.data[].__name__'

Parameters: match[] (required), start, end, limit

Metric Metadata

# Search by metric name keyword
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/metadata?metric=http_request&limit=10" | jq .

Parameters: metric (search keyword), limit.

Alerts and Rules

# All firing/pending alerts
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/alerts" | jq '.data.alerts[]'

# All alerting and recording rules
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/rules" | jq '.data.groups[]'

Instance Diagnostics

# TSDB cardinality stats
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/status/tsdb" | jq .

# Currently executing queries
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/status/active_queries" | jq .

# Most frequent/slowest queries
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/status/top_queries?topN=10" | jq .

# Version info
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/status/buildinfo" | jq .

Export Raw Data

# Export raw samples as JSON lines (one JSON object per line)
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  --data-urlencode 'match[]=http_requests_total' \
  -d 'start=2026-03-07T00:00:00Z' -d 'end=2026-03-07T12:00:00Z' \
  "$VM_METRICS_URL/api/v1/export" | head -5

# Export with reduced memory usage (for large exports)
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  --data-urlencode 'match[]={namespace="myapp"}' \
  -d 'start=2026-03-07T00:00:00Z' -d 'end=2026-03-07T01:00:00Z' \
  -d 'reduce_mem_usage=1' \
  "$VM_METRICS_URL/api/v1/export" > export.jsonl

Parameters: match[] (required), start, end, reduce_mem_usage. Output is JSON lines (not wrapped in a standard API response).

Each line: {"metric":{"__name__":"...","label":"value"},"values":[...],"timestamps":[...]}

Also available: /api/v1/export/csv (CSV format), /api/v1/export/native (binary, for import via /api/v1/import/native).

Series Count

# Total number of active time series
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/series/count" | jq .

Note: can be slow on large databases and may return slightly inflated values.

Metric Usage Statistics

# Which metrics are being queried, and how often
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/status/metric_names_stats?limit=20" | jq .

# Metrics queried <= N times (find unused/rarely-used metrics)
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/status/metric_names_stats?limit=50&le=1" | jq .

# Filter by metric name pattern
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/status/metric_names_stats?match_pattern=vm_&limit=20" | jq .

Parameters: limit (max results), le (max query count threshold — find metrics queried at most N times), match_pattern (metric name prefix filter).

Flags (Non-Default)

# View non-default runtime flags (returns plain text, one flag per line — NOT JSON)
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "${VM_METRICS_URL%%/select/*}/flags"

Root-level endpoint. Returns plain text (not JSON), one flag per line. Shows only flags that differ from defaults — useful for debugging configuration.

Query Tools

These are ROOT-level endpoints, NOT under /api/v1/. On cluster mode, strip the /select/0/prometheus prefix to get the base host.

# Expand WITH expressions (returns text, not JSON)
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "${VM_METRICS_URL%%/select/*}/expand-with-exprs?query=WITH(x=up)x"

# Prettify MetricsQL query (returns JSON)
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "${VM_METRICS_URL%%/select/*}/prettify-query?query=rate(x[5m])" | jq .

Note: ${VM_METRICS_URL%%/select/*} strips everything from /select onward, yielding the base host. On single-node (local), VM_METRICS_URL has no /select prefix so it returns the URL unchanged.

Config Debug Tools

Root-level endpoints for debugging relabeling, downsampling, and retention configurations. These accept POST with form data. Primarily used via VMUI but accessible via curl.

# Debug metric relabeling rules — test how a metric is transformed
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  -d 'metric=foo{bar="baz"}' \
  -d 'relabel_config=- target_label: cluster
  replacement: dev' \
  "${VM_METRICS_URL%%/select/*}/metric-relabel-debug"

# Debug downsampling filters — test which downsampling rules match a metric
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  -d 'metric=foo{bar="baz"}' \
  -d 'downsampling_period=30d:5m,180d:1h' \
  "${VM_METRICS_URL%%/select/*}/downsampling-filters-debug"

# Debug retention filters — test which retention policy applies to a metric
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  -d 'metric=foo{bar="baz"}' \
  -d 'retention_period=2y,{env="dev"}:30d' \
  "${VM_METRICS_URL%%/select/*}/retention-filters-debug"

These return HTML by default. For programmatic use, the VMUI at ${VM_METRICS_URL%%/select/*}/vmui/#/relabeling provides an interactive interface.

Timestamp Format

All times accept RFC3339 (2026-03-07T09:00:00Z) or Unix seconds (1709769600). Default time for instant queries is "now". Default end for range queries is "now".

Response Parsing (jq)

# Extract metric values from instant query
... | jq '.data.result[] | {metric: .metric.__name__, value: .value[1]}'

# Extract time series from range query
... | jq '.data.result[] | {metric: .metric, values: [.values[] | {time: .[0], value: .[1]}]}'

# List metric names from series response
... | jq '[.data[] | .__name__] | unique'

# Count alerts by state
... | jq '.data.alerts | group_by(.state) | map({state: .[0].state, count: length})'

# Top series by cardinality from tsdb status
... | jq '.data.seriesCountByMetricName[:10]'

Common Patterns

# Check if a metric exists
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  "$VM_METRICS_URL/api/v1/query?query=count({__name__=~\"http_request.*\"})" | jq '.data.result[].value[1]'

# Get all namespaces with active pods
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  --data-urlencode 'match[]={__name__="kube_pod_info"}' \
  "$VM_METRICS_URL/api/v1/label/namespace/values" | jq '.data[]'

# Rate of errors over last hour
curl -q --config "${VM_CURL_CONFIG:-/dev/null}" -s \
  --data-urlencode 'query=sum(rate(http_requests_total{code=~"5.."}[5m])) by (namespace)' \
  "$VM_METRICS_URL/api/v1/query" | jq '.data.result[] | {ns: .metric.namespace, rate: .value[1]}'

Environment Switching

# Check current environment
echo "VM_METRICS_URL: $VM_METRICS_URL"
echo "VM_CURL_CONFIG: ${VM_CURL_CONFIG:-(unset - no auth)}"

Important Notes

  • POST endpoints accept application/x-www-form-urlencoded (use --data-urlencode for query params with special chars)
  • match[] parameter requires the [] suffix — match alone won't work
  • For metric metadata, raw API uses metric param
  • label_values uses a path parameter: /api/v1/label/{label_name}/values
  • expand-with-exprs, prettify-query, flags, metric-relabel-debug, downsampling-filters-debug, and retention-filters-debug are root-level paths, not under /api/v1/
  • Export endpoint (/api/v1/export) returns JSON lines (one object per line), not standard API response JSON
  • Debug endpoints (metric-relabel-debug, etc.) return HTML by default — best used via VMUI for interactive debugging
  • For full endpoint details, parameters, and response formats, see references/api-reference.md

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.