k6 Performance Testing Skills
Skills for k6 performance testing with interactive planning, executor selection, script validation, and multi-environment configuration.
🎯 Overview
These skills enable Claude to plan, generate, and validate k6 performance testing scripts following official best practices. They provide deterministic recommendations, interactive parameter resolution, and comprehensive support for HTTP, gRPC, and browser testing protocols.
Key Features
- 🎯 Deterministic Recommendations: Same inputs always produce identical outputs
- 💬 Interactive Clarification: Adaptive question flow for target, scenario, SLA, protocol, and critical gaps
- 📊 Protocol Support: HTTP, gRPC, and k6/browser
- ✅ Best Practices: Follows official k6.io documentation patterns
- 🔍 Validation: Built-in script validation and quality checks
- 📝 Self-Contained Skills: All logic embedded in SKILL.md files
- 🏗️ Standards Compliant: Follows official Claude Skills specification
- 🔐 Auth-Aware Planning: Discovery gate for auth mechanism before executable output
- 🧭 Method-Aware HTTP Planning: Explicit HTTP method confirmation when required
- 🖥️ Dashboard Guidance: Deterministic recommendation to enable/disable k6 web dashboard
⚠️ Architecture Notes
This repository is aligned with the official Claude Skills specification.
Architecture:
- ✅ Self-contained SKILL.md files with all logic embedded
- ✅ Per-skill references directories for focused guidance
- ✅ Simplified plugin.json with only standard fields (auto-discovery of skills)
- ✅ Three core skills: k6-plan, k6-builder, k6-validate
- ✅ Deterministic recommendations - same inputs always produce identical outputs
🚀 Quick Start
Installation
Via skills.sh (Recommended)
npx skills add charlyautomatiza/grafana-k6-plugin
Manual Installation (Claude Desktop)
Clone to a location where Claude can auto-discover:
git clone https://github.com/charlyautomatiza/grafana-k6-plugin.git ~/claude-skills/grafana-k6
# Claude will auto-detect skills in this directory
Verification
List available skills:
npx skills list charlyautomatiza/grafana-k6-plugin
First Test
In Claude, try:
/k6-plan scenario=load target=https://httpbin.org/get sla=p95<400ms
You should get a complete, actionable test plan with recommendations.
Note: protocol is resolved in the interactive flow when needed. If omitted, the skill can ask you to confirm whether the test should use http, grpc, or browser.
🚀 First 5 Minutes
Copy and paste these prompts directly in Claude. You do not need to memorize command syntax to get started.
Plan first (/k6-plan)
I need a practical load test plan for my users API. Keep p95 below 500ms and errors below 1%.
Help me design a stress test to find the breaking point of my service, ramping safely to high load.
I need a performance plan for a gRPC service with p99 under 200ms.
Create a browser-focused performance test plan for my checkout journey and tell me the next best step.
Build runnable artifacts (/k6-builder)
Generate a runnable k6 script for my product API with thresholds aligned to p95<400ms and error<1%.
Build a multi-environment setup (dev, staging, prod) with safe env-variable placeholders.
Create runnable browser scenario artifacts for login and checkout, with clear assumptions.
