Public skills for working with Grafana, Prometheus, Loki, Tempo, Pyroscope, k6, and the broader LGTM observability stack. Compatible with Claude Code, Cursor, Codex, and any tool supporting the Agent Skills open standard.
Any tool following the Agent Skills standard can install directly:
npx skills add grafana/skills# Add this marketplace
claude plugin marketplace add grafana/skills
# Install the plugin(s) you want
claude plugin install grafana-plugins@grafana-skills
claude plugin install grafana-core@grafana-skills
claude plugin install grafana-cloud@grafana-skills
claude plugin install grafana-lgtm@grafana-skills
claude plugin install grafana-app-sdk@grafana-skills
claude plugin install grafana-k6@grafana-skillsInstall via the npx skills command shown above:
npx skills add grafana/skillsThis writes skills into .cursor/skills/ in your project so Cursor's agent can load them.
Cursor's built-in Add Rule → Remote Rule (GitHub) flow is not compatible with this repository. That importer only accepts Cursor Project Rules (
.mdcfiles under.cursor/rules/), and the Grafana Skills repo follows the Agent Skills standard instead (SKILL.mdfiles underskills/).
Skills are discovered automatically via the .agents-plugin/marketplace.json manifest. No manual setup needed — Codex loads matching skills based on your task context.
To install manually into a repo's .agents/skills/ directory:
npx skills add grafana/skillsSkills are organized into plugin groups. All skill files live under skills/<plugin-name>/.
Core Grafana concepts — dashboards, visualization, PromQL, alerting, and telemetry collection.
| Skill | Description |
|---|---|
| grafana-oss | Grafana OSS core — dashboards, data sources, provisioning, RBAC, and server configuration |
| dashboarding | Create and organise Grafana dashboards — panels, variables, transformations, and thresholds |
| promql | Write, validate, and optimise PromQL queries for Prometheus and Grafana Cloud Metrics |
| alerting-irm | Grafana Alerting, Incident Response Management, and SLOs — rules, contact points, and on-call |
| alloy | Grafana Alloy OpenTelemetry collector — config language, components, and telemetry pipelines |
| beyla | Grafana Beyla eBPF auto-instrumentation for zero-code application observability |
| opentelemetry | OpenTelemetry with the Grafana stack — SDK instrumentation, OTLP, and collectors |
Grafana Cloud — fleet management, cloud integrations, cost optimization, and AI agent connectivity.
| Skill | Description |
|---|---|
| admin | Grafana Cloud account management — organizations, stacks, RBAC, SSO, and service accounts |
| send-data | Send telemetry to Grafana Cloud — metrics, logs, traces, and profiles via Alloy or SDKs |
| fleet-management | Manage Grafana Alloy collector fleets with remote configuration and OpAMP |
| cloud-integrations | Connect AWS, Azure, and other cloud providers to Grafana Cloud |
| infrastructure | Infrastructure monitoring — Kubernetes, host/container metrics, and cloud integrations |
| app-observability | Application Observability (APM), Frontend Observability (Faro), and AI Observability |
| database-observability | Query-level performance insights for MySQL and PostgreSQL |
| adaptive-metrics | Reduce metrics cost with Adaptive Metrics aggregation rules and cardinality management |
| cost-management | Grafana Cloud cost monitoring, attribution, usage alerts, and optimization |
| dpm-finder | Identify Prometheus metrics driving high Data Points per Minute (DPM) |
| loki-label-analyzer | Evaluate and improve Loki label strategy using cardinality, query patterns, and label hygiene |
| prometheus-label-strategy | Audit and design Prometheus label schemas — cardinality, histograms, source-side prevention |
| prometheus-cardinality-troubleshooter | Diagnose live Prometheus cardinality issues — slow queries, OOMs, high Active Series bills |
| oncall-irm | Grafana OnCall and IRM — alert routing, escalation chains, and incident lifecycle |
| ml-ai | AI/ML features — Grafana Assistant, Dynamic Alerting, Sift, Knowledge Graph, and LLM Plugin |
| assistant-mcp | Connect AI coding agents (Claude Code, Cursor, Codex) to Grafana Cloud via MCP |
| private-connectivity | Private network connectivity — AWS PrivateLink, Azure Private Link, GCP Private Service Connect |
| testing | Synthetic Monitoring, k6 Cloud load testing, and Frontend Observability |
Open-source LGTM observability stack — Loki, Tempo, Prometheus/Mimir, and Pyroscope.
| Skill | Description |
|---|---|
| loki | Log aggregation with Grafana Loki — LogQL queries, pipelines, and architecture |
| tempo | Distributed tracing with Grafana Tempo — TraceQL, service graphs, and correlations |
| prometheus | Metrics with Prometheus — PromQL, alerting, recording rules, and Mimir |
| mimir | Scalable long-term metrics storage with Grafana Mimir — architecture and operations |
| pyroscope | Continuous profiling with Grafana Pyroscope — flame graphs, diff views, and language support |
Skills for building Grafana plugins — bundle optimisation, code splitting, React 19 migration, and the @grafana/scenes framework.
| Skill | Description |
|---|---|
| plugin-bundle-size | Optimise Grafana app plugin bundle size using React.lazy, Suspense, and webpack code splitting |
| react-19-plugin-migration | Migrate a Grafana plugin to React 19 compatibility ahead of Grafana 13 |
| grafana-scenes | Build Grafana plugin pages using the @grafana/scenes framework |
Skills for building apps on the Grafana App Platform using grafana-app-sdk.
| Skill | Description |
|---|---|
| app-sdk-concepts | Project init, deployment modes (standalone operator, grafana/apps, frontend-only), and workflow |
| cue-kind-definition | Author CUE kind definitions — schemas, versioning, spec vs status, codegen config |
| reconciler-logic | Implement async reconciler and watcher business logic |
| admission-control | Write validation and mutation admission handlers |
Skills for working with k6 open-source load testing.
| Skill | Description |
|---|---|
| k6 | k6 performance and load testing — test scripts, executors, thresholds, scenarios, and k6 Cloud execution |
| k6-docs | Write or review k6 documentation across TypeScript types, user docs, and release notes |
We welcome contributions! See CONTRIBUTING.md for how to add a new skill.
Quick start:
- Copy
template/SKILL.mdtoskills/<plugin-name>/<skill-name>/SKILL.md - Fill in the frontmatter and content
- Run
./scripts/lint-skills.sh skillsto validate - Open a PR
grafana-skills/
├── .claude-plugin/marketplace.json # Claude Code marketplace manifest
├── .cursor-plugin/marketplace.json # Cursor marketplace manifest (identical)
├── .agents-plugin/marketplace.json # Codex marketplace manifest (identical)
├── skill-registry.json # Machine-readable skill manifest
├── skills/ # All skills, grouped by plugin
│ ├── grafana-core/
│ ├── grafana-cloud/
│ ├── grafana-lgtm/
│ ├── grafana-plugins/
│ ├── grafana-app-sdk/
│ └── grafana-k6/
├── template/SKILL.md # Starter template for new skills
├── scripts/lint-skills.sh # Local skill validation
└── .github/workflows/ # CI validation
Apache License 2.0 - see LICENSE.