anup4khandelwal/codex-stack/qa
Trust tier 2
- community
Source check pending
Checking that the SKILL.md file is still in the repository.
Run browser flows and snapshot checks, then score release readiness.
Install
npx vskill@latest install anup4khandelwal/codex-stack/qaSKILL.md
/qa
Use this mode when the agent should validate a product surface, not just inspect code.
Objective
Run repeatable browser verification, collect evidence, and turn it into a ship/no-ship QA report.
Workflow
- Choose the target URL and named browser session.
- Import
--session-bundle <path>first when the target requires authenticated state. - Use
--mode diff-awareplus--base-refwhen the goal is to probe the pages changed by the current diff from a preview base URL. - Run one or more checked-in flows when behavior needs end-to-end validation.
- Add
--a11ywhen accessibility regressions matter, and scope it with--a11y-scopeplus--a11y-impactwhen needed. - Add
--perfwith--perf-budgetand--perf-wait-mswhen user-perceived performance needs to be enforced. - Capture or compare a snapshot when visual/text drift matters.
- Score the result with categorized findings, severity, evidence, and a recommendation.
- Load
.codex-stack/baseline-decisions/so known regressions can be downgraded, suppressed, or flagged for refresh explicitly. - Save the QA report under
.codex-stack/qa/, including annotated screenshot evidence when snapshot failures occur. - Refresh
.codex-stack/qa/trends.jsonand.codex-stack/qa/trends.mdso the operator can compare the latest run against prior QA history. - Publish tracked copies into
docs/qa/when the shipping workflow needs GitHub-linkable evidence. - Render
docs/qa/throughscripts/render-qa-pages.tswhen the operator needs a static review site.
CLI
Output format
- Overall QA status
- Health score
- Findings with category + severity
- Diff-aware route probe summary
- Snapshot evidence
- Accessibility and performance summaries plus artifact paths
- Regression triage summary plus applied, expired, and unresolved decisions
- Flow pass/fail summary
- Recommendation
- Trend delta vs previous runs via
.codex-stack/qa/trends.json
Guardrails
- Do not claim QA passed without runtime evidence.
- Prefer checked-in flows and named snapshots for repeatable environments.
- Use
--session-bundleinstead of embedding secrets into flows when preview or staging authentication is required. - Use
--mode diff-awareonly when the provided URL is the deploy base, not a single leaf page. - Refresh a snapshot baseline only when the UI change is intentional.
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/qa/SKILL.md
Popularity
- 7-day trend score
- 5
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/anup4khandelwal/codex-stack/qa)