qa-gate
Skill10 starsChanged 26 days agoMITView on GitHub
Run a 9-phase pre-delivery QA gate on a build and return a PASS/CONDITIONAL/FAIL verdict with a risk rating (Low/Medium/High/Critical), bucketed findings (critical issues, warnings, recommendations, nice-to-haves), an estimated fix time, and an explicit client-ready YES/NO. Uses the Playwright extension when connected for live functional, visual, and performance checks; otherwise runs the static path and marks live-only checks N/A. Trigger when the user says "qa check", "qa gate", "is this ready to ship", "client-ready check", "pre-delivery review", or "stress test the site".
6 other projects use a version of this.
Adopted by
Most recently taken up by
- GeorgeAuth26/authentalink-playbooktheir own version, first seen Nov 8, 2026
- weedmo/skillstheir own version, first seen Nov 8, 2026
- mementomori-co/maker-masterytheir own version, first seen Oct 25, 2026
- manhds82/claude-code-anyllmtheir own version, first seen Oct 13, 2026
- victoraguilarsantamariadev/qa-exploretheir own version, first seen Sep 28, 2026
- Daniellaguips/ai-agent-ostheir own version, first seen Jun 10, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- manhds82/claude-code-anyllm★ 16same name
- mementomori-co/maker-mastery★ 2same name
- victoraguilarsantamariadev/qa-explore★ 2same name
- GeorgeAuth26/authentalink-playbook★ 1same name
- weedmo/skills★ 1same name
- Daniellaguips/ai-agent-os★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository46 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- blast-promptSkill
- client-outreachSkill
- content-draftSkill
- copywritingSkill
- csv-to-reportSkill
- design-accessibilitySkill
- design-buildSkill
- design-croSkill
- design-dimensionsSkill
- design-motionSkill
- design-researchSkill
- design-system-genSkill
- design-system-persistSkill
- design-tokens-emitSkill
- design-visual-qaSkill
- html-extractSkill
- parallel-buildSkill
- portable-html-portSkill
- route-codex-reviewSkill
- route-gemini-contextSkill
- route-three-brainSkill
- seo-auditSkill
- seo-backlinksSkill
- seo-clusterSkill
- seo-competitor-pagesSkill
- seo-content-briefSkill
- seo-contentSkill
- seo-dataforseoSkill
- seo-driftSkill
- seo-ecommerceSkill
- seo-firecrawlSkill
- seo-geoSkill
- seo-googleSkill
- seo-hreflangSkill
- seo-image-auditSkill
- seo-image-genSkill
- seo-local-unifiedSkill
- seo-pageSkill
- seo-programmaticSkill
- seo-schemaSkill
- seo-sitemapSkill
- seo-strategySkill
- seo-sxoSkill
- seo-technicalSkill
The file
Files
qa-gate
Family: build-and-qa Status: Stable
Purpose
The mandatory pre-delivery gate for any client-bound build. Walks the build through 9 phases and produces a structured pass/fail report with a risk rating and an explicit YES/NO client-ready verdict. This is the gate, not the fix — it reports; remediation happens in the appropriate skills.
Triggers
- "qa check" / "qa gate"
- "is this ready to ship" / "client-ready check"
- "pre-delivery review"
- "stress test the site"
Inputs
- Build URL or local path
- Target client/audience
- Deployment target (Vercel, WordPress, GHL, etc.)
- Brief or original spec (for the completeness check)
Steps
- Load the report template
templates/qa-report-template.md. - Run the static gate in one command — it drives the plugin's own engines
(
a11y_static,tech_audit,image_audit,content_audit,schema_gen,link_graph) over every page of the build, adds the exposed-secret scan and the deployment checks (stagingDisallow: /left on, robots/sitemap/404/favicon, analytics), merges repeats across pages, applies the verdict rule, and fills the template:
(python3 "${CLAUDE_PLUGIN_ROOT}/scripts/workflow/qa_gate.py" --dir dist/ --base-url https://client.example --project "<client>" --report--file page.html --url <URL>for a single page; JSON without--report.) Then detect tooling: if the Playwright extension is available (seeextensions/playwright/), deepen the live phases below; if not, the static report already marks themN/A — needs Playwright. - Run the 9 phases, recording PASS / WARN / FAIL / N/A for each:
- Functional — forms submit, links resolve, no console JS errors.
- Visual fidelity — matches the design system; responsive at 375/768/1280;
no broken layouts. (Playwright deepens this;
design-visual-qafor diffs.) - Accessibility — dispatch the
design-accessibilityagent (axe + WCAG AA when Playwright is connected; else its bundleda11y_static.pystructural checks — contrast, alt text, heading order, labels, focus visibility). - Performance — Core Web Vitals (LCP < 2.5s, CLS < 0.1, INP < 200ms).
Dispatch the
seo-googleagent for PSI/CrUX field data when a key/MCP is connected (the free PSI/CrUX key is preferred; any paid call routes throughcost_guard.py, which fails open to the free path); else flag CWV for field verification. - Security — HTTPS, security headers, no exposed secrets in source, audit third-party scripts.
- Content completeness — no lorem ipsum, no broken images, no placeholder copy; compare against the brief for missing pieces.
- SEO baseline — dispatch the
seo-pageagent (title, meta, OG, schema, canonical, images, internal links) as a real fan-out, not a prose handoff. - Cross-device / browser — sanity at key breakpoints/browsers.
- Deployment readiness — env vars set, redirects, analytics installed, DNS.
- Apply the verdict rule. Any unresolved CRITICAL ⇒ OVERALL = FAIL and CLIENT-READY = NO. CONDITIONAL PASS only when remaining items are WARN or lower.
- Estimate fix time and bucket findings into Critical / Warnings / Recommendations / Nice-to-haves.
- Render the filled
qa-report-template.md. Never declare client-ready with an open critical issue.
Outputs
Structured QA Summary Report (from templates/qa-report-template.md): overall
status, risk rating, per-phase results, bucketed findings, estimated fix time, and
an explicit client-ready YES/NO.
Dependencies
templates/qa-report-template.md(required)scripts/workflow/qa_gate.py(required) — the static 9-phase runner (imports the SEO, content, image, schema, link-graph and a11y engines)design-accessibility(optional — adds axe/WCAG delegation; free path: static contrast/alt/heading/label/focus checks) — phase 3seo-page(optional — adds full on-page SEO pass; free path: inline title/meta/OG/canonical check) — phase 7seo-google(optional — adds PSI/CrUX field data; free path: flag CWV for field verification) — phase 4design-visual-qa(optional — adds screenshot diffs; free path: manual visual review) — phase 2/8- Playwright extension (optional — deepens phases 2/3/4/8; static path otherwise)
Notes
This is the gate, not the fix. Remediation happens in design-build,
seo-technical, etc. Pairs with blast-prompt as the build bookend. Never skip
this gate when delivering to a paying client.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

