devtools
Skill9 starsChanged 2 months agoMITView on GitHub
Frontend DevTools audit workflow — Lighthouse/CWV, a11y (axe), network/layout, when checks are MUST vs optional. Use for /test-ui deep passes, "audit performance", "cek Lighthouse", "axe", or verifying UI before ship with tooling evidence.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- vercel-labs/json-rendertheir own version, first seen Oct 29, 2026
- OHOKAYYEAH/claude-skillstheir own version, first seen Oct 13, 2026
- apache/tvm-ffitheir own version, first seen Oct 8, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- vercel-labs/json-render★ 19ksame name
- apache/tvm-ffi★ 477same name
- OHOKAYYEAH/claude-skills★ 1same 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 repository36 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .cursor/rules/design-system-enforcer.mdcCursor rule
- frontend-expertSkill
- accessibilitySkill
- anti-slop-codeSkill
- anti-slop-designSkill
- app-shellSkill
- architectureSkill
- componentsSkill
- content-designSkill
- data-fetchingSkill
- design-colorSkill
- design-directionSkill
- design-foundationsSkill
- design-surfacesSkill
- design-typographySkill
- design-vocabularySkill
- engineering-vocabularySkill
- fidelitySkill
- formsSkill
- frontend-expertSkill
- marketingSkill
- monitoringSkill
- motion-inspectionSkill
- motionSkill
- performanceSkill
- polishSkill
- prototypeSkill
- quality-loopSkill
- responsiveSkill
- seoSkill
- skill-authoringSkill
- testingSkill
- tokensSkill
- webglSkill
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 13
Use browser/tooling evidence when available — don’t invent Lighthouse scores.
- Distinctive: line 24
No runnable app / static markdown-only change
- Very specific: line 39
Rule: Never fabricate scores. If tools didn’t run → scorecard = not evaluated.
“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
FE DevTools
Overview
Use browser/tooling evidence when available — don’t invent Lighthouse scores. Pair with performance, accessibility, and testing.
When to use
/test-uior ship checks when tooling exists in the project- User asks for Lighthouse, axe, Network, Layout shift, Coverage
- Design/perf audit needs measured Performance or Accessibility scorecard rows
- Verifying no horizontal overflow / focus order with real viewport tools
When to skip
- No runnable app / static markdown-only change
- User wants theory-only advice with no environment
- Pure token rename with no UI surface
MUST vs optional
| Check | MUST when | Optional |
| Responsive viewports | Layout UI ship (320/768/1024/1440) | — |
| Keyboard / focus | Interactive forms, dialogs, drawers | — |
| axe / a11y scan | /test-ui, /design deep, or user asks | Quick source-only audit |
| Lighthouse / CWV | Perf ask, public landing, /design deep with app running | Tiny internal chrome tweak |
| Network waterfalls | Suspected overfetch / huge assets | Default CRUD form |
| Layout / CLS | Animations, fonts, images without dimensions | Static settings copy |
Rule: Never fabricate scores. If tools didn’t run → scorecard = not evaluated.
Workflow
- Confirm target — URL/route or local preview command
- Viewport — resize or device mode; note Responsive Conventions
- A11y — axe or equivalent; fix Critical/High (labels, contrast, names)
- Perf — Lighthouse or Web Vitals extension on key route; note LCP/INP/CLS risks
- Network — flag obvious huge JS/CSS/images on first load
- Report — tool · metric · value · source; hand remediations to
performance/accessibility/ build skills
Checklist
- Did not invent Lighthouse/axe numbers
- Viewport evidence for layout ships
- A11y scan when MUST (or waiver documented)
- Perf metrics only when tool ran or explicitly source-level risks
- Findings mapped to skills (not dumped as vague “optimize”)
Depth
references/devtools.md.
CWV detail: references/performance.md.
A11y: references/accessibility-checklist.md.
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.