Figma Design QaCAUTION

1255 professional Agent Skills for Claude, ChatGPT, Gemini, Cursor & Codex — PRDs, postmortems, leases, medical bills, layoffs, go-bags, new countries. Plain markdown, MIT, in Anthropic's plugin directory. Plus a typed decision layer: which skill, is it safe, ship or slip. Free in-browser or npx pm-

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: figma-design-qa
description: "Runs a pre-handoff QA checklist on a Figma design before it goes to engineering. Use when asked to QA a Figma design, do a pre-handoff check, or validate a Figma file is ready to build. Produces a structured QA report covering file hygiene, component usage, accessibility, and handoff readiness with explicit pass/fail status per item. Optimised for Opus 4.7 and newer models."
---

# Figma Design QA Skill

Runs a systematic pre-handoff QA check on a Figma design — catching issues that cause engineering back-and-forth before they become expensive.

> **Not quite this?** Use `figma-design-brief` when the design does not exist yet and needs a brief; use `figma-design-review` when you want a holistic design review rather than a check against the spec; use `figma-design-critique-pm` when you are judging product outcomes rather than implementation fidelity.

## Required Inputs

Work from what is given. If any of these is missing, still produce the complete output: fill each gap with a reasonable assumption marked *(assumed, confirm)* and list those assumptions at the end. Ask a question first only when the core subject itself is missing (there is nothing to write about).
- **Feature or screen being QA-d** (describe what has been designed)
- **Platform** (iOS / Android / Web)
- **Design system** (custom / Material / HIG / None)
- **Handoff tool** (Figma Inspect / Zeplin / Storybook / Direct link)
- **QA depth** (quick 15 min / standard 30 min / thorough 60 min)


## Programmatic Helper

Contrast ratios cannot be eyeballed. The AA line sits at 4.5:1, and `#777777`
on white is 4.478 (fails) while `#767676` is 4.54 (passes) — no amount of
looking at a screenshot separates those. Compute them:

```bash
npx --yes notugly fix "#8ab4f8" "#ffffff"     # ratio, APCA, and the nearest passing colour
npx --yes notugly onepager <url> --out review.html   # every pairing, printable
npx --yes notugly vision                      # which colours merge for colour-blind viewers
```

`notugly fix` returns the ratio, the APCA lightness contrast, and the closest
colour to the one already chosen that passes — same hue, same chroma. Paste
those numbers into the tables below rather than estimating them.

Deterministic, zero dependencies, and **no model call** — so it costs nothing to
run and gives the same answer every time.

QA is a pass/fail activity, so every contrast row needs a number and a verdict.
For a deployed build, `npx notugly check <url>` **exits non-zero** on a failure,
which makes it usable directly in CI rather than only by hand.

## Output Structure

QA Report: [Feature] | [Date] | [Platform]
**Overall status:** Ready / Minor fixes needed / Not ready

### Section 1: File Hygiene
- All layers named semantically (no "Rectangle 12")
- No unused/hidden layers in final frames
- Components from library (not detached copies)
- All text uses text styles (not manual font settings)
- All colours use styles or variables (not hex overrides)
- Frames named to match screen map
- No leftover prototype wires to wrong frames

### Section 2: Component Usage
- All buttons use library component
- All inputs use library component
- All icons from approved icon library
- No custom components that should be in library
- Variants used correctly (right size, state, type)

### Section 3: Content and Copy
- No placeholder text (Lorem ipsum) in final designs
- All copy reviewed and approved
- Realistic content used (not "User Name")
- Long text edge cases tested
- Error messages are human-readable
- Empty states have copy and CTA

### Section 4: States and Coverage
- Default, Loading, Empty, Error, Success states
- Interactive elements have hover/active (web)
- Disabled states designed where applicable

### Section 5: Accessibility
- All text meets WCAG AA contrast (4.5:1 body, 3:1 large)
- UI components meet 3:1 contrast against background
- Touch targets minimum 44x44pt iOS / 48x48dp Android
- Focus states for keyboard/switch navigation (web)
- Information not conveyed by colour alone
- Icons have text labels or accessible names annotated

### Section 6: Handoff Readiness
- Dev annotations on non-obvious interactions
- Spacing uses Auto Layout (not absolute positioning)
- Images/assets exported at correct resolutions
- Design matches approved requirements
- Link to prototype included

### Issues Found
For each fail:
**[Issue] — Blocking / Fix before handoff / Fix in next iteration**
- What: [Specific layer/screen/element]
- Fix: [Exact action needed]
- Owner: [Designer/PM/Both]

### Handoff Decision
Status, signed off by, date.

## Quality Checks
- [ ] All 6 sections completed
- [ ] Every fail has a specific description and fix action
- [ ] Blocking issues separated from minor ones
- [ ] Handoff decision is explicit

## Anti-Patterns

- [ ] Do not produce a partial QA — every checklist category must be evaluated, not just the ones that are obviously problematic
- [ ] Do not leave the handoff decision ambiguous — the output must explicitly state pass, pass with conditions, or fail
- [ ] Do not skip accessibility checks — colour contrast, tap target size, and screen reader labels are required, not optional
- [ ] Do not report issues without specifying which screen or component they appear on
- [ ] Do not approve a design if any component is detached from the library without a documented reason

## Example Trigger Phrases
- "QA this Figma design before handoff"
- "Run a pre-handoff check on [feature] design"
- "Is this Figma design ready for engineering?"
- "Do a design QA on [screen/feature]"
- "What needs fixing before we hand this off?"

CAUTIONgrade B · trust 89/100 Install with care. The audit found things worth knowing before you trust its output.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (5)

templates/pm-discovery-agent/skills/assumption-mapper/SKILL.md

templates/pm-discovery-agent/skills/discovery-interview-guide/SKILL.md

templates/pm-discovery-agent/skills/job-story-mapper/SKILL.md

templates/pm-discovery-agent/skills/user-interview-synthesis/SKILL.md

templates/pm-launch-agent/skills/content-calendar/SKILL.md

Gates applied: no_behavioural_pass.