Observability
Agent skills and MCP servers for observability that Spellwarden has checked. Each entry shows what was checked and when.
victoriametrics-operator
Reference for VictoriaMetrics Operator CRDs: what each kind does, key fields, example manifests and common mistakes.
No open critical or high findings. 1 open medium or low finding.
Curated
prometheus
Guidance for writing Prometheus scrape configs, recording and alert rules, and PromQL, with rules to keep label cardinality bounded.
No open critical or high findings. 1 open medium or low finding.
Curated
Axiom
Lists Axiom datasets and schemas and runs APL queries against them, with prompts for log exploration, anomaly hunting and monitoring.
No open critical or high findings.
azure_managed_grafana
Guides an agent through Azure Managed Grafana MCP tools to query metrics, logs, traces, KQL, SQL and Resource Graph and find dashboards.
No open critical or high findings. 2 open medium or low findings.
Curated
observability
Guidance for designing and reviewing production logging, metrics, SLOs, tracing, alerts and telemetry pipelines, with reference checklists.
No open critical or high findings. 1 open medium or low finding.
Curated
prometheus-cardinality-troubleshooter
Diagnoses live Prometheus cardinality problems using tsdb status endpoints, PromQL drill-downs, churn checks and safe whole-metric drop configs.
No open critical or high findings. 2 open medium or low findings.
Curated
prometheus-label-strategy
Guidance for auditing and designing Prometheus label schemas to limit cardinality at the source, with example relabel and recording rule configs.
No open critical or high findings. 1 open medium or low finding.
Curated
prometheus
PromQL reference with alerting, Alertmanager routing and recording rule examples, plus promtool and amtool validation commands
No open critical or high findings. 2 open medium or low findings.
Curated
promql
Query patterns and checks for PromQL on Prometheus, Mimir and Grafana Cloud: rates, histograms, SLO burn rates, recording rules, cardinality
No open critical or high findings. 1 open medium or low finding.
Curated
Honeycomb
Hosted MCP endpoint for querying Honeycomb traces, events, metrics, SLOs, triggers and boards; requires OAuth sign-in.
No open critical or high findings.
incident-postmortem
Guides a blameless incident postmortem (timeline, root cause, owned action items) with a local script that sorts events and computes TTD, TTM and TTR.
No open critical or high findings. 2 open medium or low findings.
Curated
observability-setup
Adds OpenTelemetry logs, metrics and traces to a service, defines SLOs and generates Prometheus burn-rate alerts with two offline scripts.
No open critical or high findings. 3 open medium or low findings.
Curated
New Relic
Queries New Relic observability data (metrics, logs, traces, entities) over a hosted MCP endpoint; requires a New Relic OAuth sign-in.
No open critical or high findings. 1 open medium or low finding.
Rootly
Remote MCP server for Rootly incident management: incidents, on-call schedules and incident analysis. Requires OAuth sign-in.
No open critical or high findings. 1 open medium or low finding.
Sentry
Sentry's hosted MCP server for looking up errors, issues and debugging data. Requires an OAuth sign-in before it lists its tools.
No open critical or high findings.