Argus — Autonomous UI Auditor (v2.7). 92 skills covering layout, typography, accessibility, forms, navigation, data controls, content review, RTL, dark mode, CSS compat, video, and AI vision review. Deterministic Playwright runner executes all 57 passive probes (no skill-collapse, no fabricated issueTypes), evidence-based coverage gate, console/network DevTools-panel screenshots, red annotated boxes with issue labels. Auto-installs Playwright + MCP, streams findings to disk, files Azure DevOps bugs with annotated evidence.
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Argus QA readiness wizard; safe alias for qa-ready that avoids old qa-agent command collisions
Argus QA first-run setup wizard; configures MCP, app settings, and Azure DevOps credentials
Deduplicates issues.jsonl and files Azure DevOps Bug work items with annotated screenshots attached
Parallel cell-execution subagent. Receives a chunk of audit cells, opens its own browser tab, runs them sequentially within the tab, writes findings to JSONL, returns. Spawned in parallel by qa-argus Step 5.1 when resolvedConfig.workers > 1. Honors the workers setting dynamically: workers=4 spawns 4 of these in parallel, workers=8 spawns 8.
Reads issues.jsonl after the audit, writes coverage-report.json, and prints a terminal summary. Does NOT write a markdown file.
Autonomous QA engineer for Claude Code. Crawls every route × viewport × browser of your app, detects defects, and files Azure DevOps bugs with screenshots — automatically.
Runs a single command — /argus-qa:qa-argus — and gets you:
Works on any frontend framework (Angular, React, Vue, Next.js, Nuxt, SvelteKit, Remix, Astro, plain HTML — anything Playwright can navigate to).
After installing the plugin, in your project root:
/argus-qa:qa-argus
On first run, a setup wizard asks for:
The wizard writes .claude/automation.config.json (non-sensitive config) and .claude/secrets.json (PAT + passwords, gitignored). Then the audit runs.
On every subsequent run, the agent just runs the audit using your saved config.
| Command | Purpose |
|---|---|
/argus-qa:qa-argus | Full audit — all apps, all routes, all viewports, all browsers |
/argus-qa:qa-argus --app <name> | Audit one app only |
/argus-qa:qa-argus --route <path> | Audit one specific route |
/argus-qa:qa-argus --device mobile | Audit one viewport only |
/argus-qa:qa-argus --browser chromium | Use only one browser |
/argus-qa:qa-argus --dry-run | Detect issues but do not file ADO bugs |
/argus-qa:qa-argus --headless | Run browsers invisibly (faster) |
/argus-qa:qa-argus --no-vision | Skip the AI vision review |
/argus-qa:qa-argus --resume <runId> | Resume an audit that crashed mid-run (skips completed cells) |
/argus-qa:qa-argus-setup | Re-run the setup wizard |
0. Step 0 auto-invokes qa-argus-setup if .claude/automation.config.json is missing
1. /argus-qa:qa-argus-setup wizard → writes .claude/automation.config.json on first run
2. /argus-qa:qa-preflight → URL health, auto-installs Playwright MCP if missing,
auto-installs node_modules, verifies server, secrets
3. /argus-qa:qa-route-discovery → crawls routes via Playwright MCP (Sonnet for SPA judgment)
4. /argus-qa:qa-phase-strategy → orders routes into 3 phases (public → login → auth-gated)
5. Per-cell execution → orchestrator calls Playwright MCP per cell, dispatches each
enabled skill on its declared model (Haiku for mechanical
probes, Sonnet for judgment, vision review, content review)
6. /argus-qa:qa-vision-review → AI vision pass on critical-route screenshots (Sonnet vision)
7. /argus-qa:qa-bug-filer → deduplicates and files ADO bugs with screenshots (Sonnet)
8. /argus-qa:qa-coverage-report → prints a summary
9. qa-state.json updated → tracks chronic issues across runs
Each skill is self-contained. Browser operations happen through Playwright MCP tools (or inline-Bash fallback). No permanent Playwright runner script exists.
🔍 QA Agent — Run qa-20260514-rp02
Phase 1 — Public routes (4 cells) ✓
Phase 2 — Login route (1 cell) ✓ session established
Phase 3 — Auth-gated routes (7 cells) ✓
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimnpx claudepluginhub luqman-ud-din/blackbox-qa-agent --plugin argus-qaComprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
A growing collection of Claude-compatible academic workflow bundles. Covers scientific figures, manuscript writing and polishing, reviewer assessment, citation retrieval, data availability, paper reading, literature search, response letters, paper-to-PPTX conversion, and evidence-grounded Chinese invention patent drafting. Rules are organized as reusable skill folders with explicit workflows and quality checks.
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
Develop, test, build, and deploy Godot 4.x games with Claude Code. Includes GdUnit4 testing, web/desktop exports, CI/CD pipelines, and deployment to Vercel/GitHub Pages/itch.io.
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.