Introduction

dq-awesomeqa is a plugin for coding agents that support plugins. It guides QA engineers through the full Software Testing Life Cycle (STLC) across four domains:

Domain	What it tests
🌐 UI	End-to-end browser flows, visual regression
🔌 API	REST / GraphQL contract, schema validation, edge cases
♿ Accessibility	WCAG 2.1/2.2 AA/AAA compliance
⚡ Performance	Load, stress, and soak scenarios

The plugin orchestrates three specialised tools under the hood - you never invoke them directly. dq-awesomeqa acts as the QA Expert who knows when and how to use each one.

---

Core philosophy

QA Expertise, not just tooling. Every skill explains why it asks what it asks, interprets findings in professional QA language, and closes with clear recommended next steps.

Four principles drive every design decision:

1. Config-first - All skills read dq-qa.config.json before acting. Missing config triggers /qa-onboard automatically.
2. Domain isolation - Each domain skill can run independently. A "Performance only" cycle is a first-class path.
3. Tools do the heavy lifting - Code generation delegates to the underlying CLIs. Skills orchestrate and refine, never generate from scratch.
4. Separate reports, unified summary - Each domain keeps its native report. /qa-report aggregates them into one dashboard.

---

STLC at a glance

/qa-init
  │
  ├─ Phase 1 - Setup
  │     /qa-setup    → install CLIs (once per machine)
  │     /qa-onboard  → configure project (once per project)
  │
  ├─ Phase 2 - Planning
  │     /qa-requirement → gather requirements → requirements/ folder
  │     /qa-impact      → record changes (subsequent cycles)
  │     /qa-plan        → derive test strategy → qa-plan.md
  │
  ├─ Phase 3 - Design  (only in-scope domains)
  │     /qa-api      → API test plan
  │     /qa-ui       → UI interaction YAML
  │     /qa-a11y     → WCAG audit steps
  │     /qa-perf     → load test config
  │     /qa-coverage → design gap check
  │
  ├─ Phase 4 - Execution
  │     /qa-exec     → run all in-scope tests
  │
  └─ Phase 5 - Closure
        /qa-triage   → categorize failures, ship/no-ship verdict
        /qa-coverage → release readiness gate
        /qa-report   → consolidated qa-summary.md

---

Built-in feedback loops

Two feedback loops are baked into the journey so no failure falls through the cracks:

• Test failures → /qa-triage → fix → re-execute
• Requirement changes → /qa-impact → updated plan → re-design

---

Prerequisites

Requirement	Version
A coding agent with plugin support (Claude Code, GitHub Copilot, Gemini CLI)	Latest
Node.js	≥ 18
.NET SDK (for performance testing only)	≥ 8
jq (recommended)	Any

---

Next steps

• New to the plugin? → Installation
• Ready to run? → Quick Start
• Want the full picture? → Architecture Overview