Grafana Skills
The official Grafana skill set: PromQL, Loki, Tempo, Alloy, k6, dashboards and alerting for Claude Code, Cursor and Codex
- 289 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- The Grafana Cloud skills lead the agent through admin APIs and CLIs, which needs tokens of your account
- The k6 load skills prepare tests that put real load on the site you name
- The dependency and package manager skills suggest changing npm, yarn and pnpm configuration in your project
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/grafana-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The Grafana Cloud skills lead the agent through admin APIs and CLIs, which needs tokens of your account
- The k6 load skills prepare tests that put real load on the site you name
- The dependency and package manager skills suggest changing npm, yarn and pnpm configuration in your project
This is third-party code. Review the repository files before installing.
What it does
Grafana Skills is a public set of skills for the Grafana, Prometheus, Loki, Tempo, Pyroscope and k6 stack. The skills are grouped into seven plugins: grafana-core (dashboards, PromQL, alerting and IRM, Alloy, Beyla, OpenTelemetry), grafana-cloud (administration, sending telemetry, Alloy fleet management, metrics cost, Frontend Observability, OnCall), grafana-lgtm (Loki, Tempo, Prometheus, Mimir, Pyroscope), grafana-plugins, grafana-app-sdk, grafana-k6 and grafana-datasources. Some skills teach the agent to write and read PromQL, LogQL and TraceQL queries. Others walk through tasks: design a Prometheus label schema, find the cause of high cardinality, diagnose a Frontend Observability session from a gcx dump, write a k6 script through a generate, validate, review loop. For authors of Grafana plugins there are skills on bundle size, the React 19 migration and the @grafana/scenes framework. Every skill sits in skills/<plugin>/ in the Agent Skills format, and many have a references folder beside them.
Who it is for. For SREs, DevOps engineers and developers who run monitoring on Grafana and write queries, dashboards, alerts and load tests with an agent.
Good fit when
- You need to write or speed up a PromQL query and the agent should know the difference between rate, irate and increase
- Your metrics bill keeps growing and you need to find high-cardinality series and Adaptive Metrics rules
- You want to build a dashboard with panels, variables, transformations and thresholds
- You want to write a k6 load test and review it through a working loop
- You develop a Grafana plugin and want to shrink the bundle or move to React 19
Not a fit when
- The agent has to run queries against a live Grafana: an MCP server does that, while the skills supply knowledge and workflow
- You do not use Grafana, Prometheus or related products: the whole set is about that stack
- You want a single skill from outside your area: pick it with --skill instead of installing the whole set
- You want to add the repository through Remote Rule in Cursor: that importer accepts only .mdc files and the format here is different
Example request
Rewrite this PromQL query so it does not overload Prometheus and explain why rate beats irate hereLimitations
The skills give the agent instructions and references. To work with live data you need access to your Grafana or Grafana Cloud and companion tools such as gcx or an MCP server. The built-in Remote Rule import in Cursor is incompatible with this repository, so install through npx skills.
The set keeps growing, so check the README tables and the skill-registry.json file for the current list of skills.
How to disable. Delete the skill folders from the project .claude/skills/, .cursor/skills/ or .agents/skills/. Claude Code plugins come off with claude plugin uninstall, for example claude plugin uninstall grafana-core@grafana-skills.
FAQ
How do I install the Grafana skills in Claude Code?
Run claude plugin marketplace add grafana/skills, then install the plugins you need, for example claude plugin install grafana-core@grafana-skills. The plugins are grafana-core, grafana-cloud, grafana-lgtm, grafana-plugins, grafana-app-sdk, grafana-k6 and grafana-datasources.
How do I install the Grafana skills in Cursor?
Run npx skills add grafana/skills: the command writes the skills into the project .cursor/skills/. Remote Rule in Cursor settings does not fit this repository because it accepts only .mdc files.
Can I install one skill instead of the whole set?
Yes, add the --skill flag with a name. The example from the set: npx skills add grafana/skills --skill frontend-session-rca installs only the Frontend Observability session analysis.
How are the Grafana skills different from Grafana MCP?
An MCP server gives the agent tools to query a live Grafana, while the skills explain how to write queries, design labels and run workflows. The assistant-mcp skill in this set describes connecting Claude Code, Cursor and Codex to Grafana Cloud through MCP.
Which k6 tasks does the set cover?
The grafana-k6 plugin has seven skills: generating and validating scripts, load testing a whole website, managing runs in Grafana Cloud k6, investigating a specific run, trend analysis, maintaining existing tests and k6 documentation.
Discussion
No questions or reviews yet.
Related
SkillsKubeShark skill: the agent finds the failure mode first, then writes manifests, Helm and Kustomize to official practice with a rollback plan
Docker Skills
docker-skills
Official Docker skills for agents: Dockerfile and builds, Compose, Sandboxes, Docker Agent and a skill that confirms irreversible actions
Cloudflare's official agent skills: Workers, Durable Objects, wrangler, Pages, storage and an MCP server
A large base of DevOps, security, infrastructure and compliance skills for AI agents
31 DevOps skills for Claude Code and Codex: config generators, validators and a Kubernetes debugger for Terraform, Helm, CI and Docker
Agent Skills for Google products
Agent Skills for Google products and technologies
Official Google skill collection for working with Google Cloud, BigQuery, GKE, ads and analytics from an agent
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Grafana Skills (this entry) | yes | 289 | ||
| Grafana MCP Server | yes | API key | 3,587 | an MCP server that runs queries against a live Grafana, while this set holds knowledge and workflows |
| Prometheus MCP Server | no | not required | 519 | an MCP server for querying Prometheus from an agent, without teaching skills for the stack |
README in short
The description says the skills fit Claude Code, Cursor, Codex and other tools that follow the Agent Skills standard. It then gives install commands for each, followed by tables of skills across the seven plugins with one line per skill. It closes with short rules for adding a new skill and a map of the repository folders with marketplace manifests for Claude Code, Cursor and Codex.
README badge
Original SKILL.md
---
name: promql
license: Apache-2.0
description: Write, validate, and optimize PromQL for Prometheus / Grafana Mimir / Grafana Cloud Metrics. Covers `rate` vs `irate` vs `increase`, label matchers and regex, `sum / avg / topk / by / without` aggregation, classic + native `histogram_quantile`, ratios with divide-by-zero guards, `absent` / `changes` for staleness, time offsets and `predict_linear`, recording-rule naming, SLO + burn-rate math, and a cardinality-hunting playbook. Use when writing a metric query, fixing wrong p95s, building an error-budget alert, debugging "query is slow", finding the noisy label that blew up cardinality, or migrating a dashboard query to a recording rule — even when the user says "calculate the error rate", "p99 latency", "sum by service", "why is this query slow", or "what's filling Mimir" without naming PromQL.
---
# PromQL Query Patterns
> **Docs**: https://prometheus.io/docs/prometheus/latest/querying/basics/
PromQL returns either an **instant vector**, a **range vector**, or a **scalar**.
**Golden rule:** `rate()` / `increase()` require a range vector ≥ 4× the scrape interval. 60s scrape → use `[5m]` minimum.
## Prerequisites
- A Prometheus / Mimir / Grafana Cloud endpoint to query (`/api/v1/query` or via Grafana Explore)
- The PromQL pattern library in [`references/patterns.md`](references/patterns.md)
## Common Workflows
### 1. Write + validate a query
```bash
# 0. Point at your Prometheus/Mimir. For Grafana Cloud, use the metrics endpoint
# and add basic auth (-u "<metrics_user>:<token>") to each curl below.
PROM=http://localhost:9090 # or https://prometheus-prod-XX.grafana.net/api/prom
# 1. Sketch the query — for "5xx error rate per service":
EXPR='sum(rate(http_requests_total{status_code=~"5.."}[5m])) by (service)'
# 2. Validate syntax + that the metric/labels exist
curl -sG --data-urlencode "query=${EXPR}" \
"$PROM/api/v1/query" | jq '.status, (.data.result|length)'
# Expect: "success" and result count > 0. If 0 — check label spelling and scrape activity:
curl -sG --data-urlencode "match[]=http_requests_total" "$PROM/api/v1/series" | jq '.data | length'
# 3. Sanity-check the magnitude — open Grafana Explore, paste the expr,
# confirm the values look right against a known ground truth (k6 run, log count, etc.)
```
### 2. Common patterns to copy
**Per-status request rate** (aggregate AFTER rate):