AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

design-qa

Use when a UI screen, page or component is about to be called done, handed off or shipped, or when generated UI looks generic or half-finished. Gates the rendered result, not the design file - finds controls that do nothing, placeholder copy, removed focus rings, missing loading/empty/error states, low contrast, small targets, hidden focus and broken reflow measured in the real browser, and flags template defaults without blocking on taste.

Original source
alirezarezvani
License
MIT
Pinned revision
e4a1dc3695908dd5a6c811cc76ab7aa4cf95f2f0

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

This skill also requires local files or dependencies. Remote reading alone may not be enough.

Ask your assistant

Use Agentic Memory skills to read alirezarezvani--design-qa at revision e4a1dc3695908dd5a6c811cc76ab7aa4cf95f2f0, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 5

assets/finish_signoff.md

UI finish sign-off

Fill this in from the script runs before a screen is called done. Paste verdict lines, not summaries. A BLOCKED line means the screen is not done.

Screen / flow: Build or commit: URL and viewports captured: 320 × ___ and ___ × ___ Brief or design file: Reviewer (a person, not the agent):

1. Finish lint (finish_lint.py)

  • Verdict line:
  • Blockers fixed:
  • Suppressions added, each with its reason:
  • Template traits and generic risk:
  • If generic risk is MEDIUM or HIGH, does the brief ask for this look? (yes, quote it / no, and what was changed)

2. State matrix (state_matrix_checker.py)

  • Verdict line and coverage %:
  • States marked NA and why:
  • Product-specific states added with --require:

3. Rendered audit (rendered_audit.py), once per viewport and per key state

| Snapshot | Viewport | State | Verdict | Blockers | Waivers | |---|---|---|---|---|---| | | 320 | populated | | | | | | desktop | populated | | | | | | 320 | empty | | | | | | desktop | error | | | |

  • Focus data came from a real keyboard pass (capture_snapshot.mjs): yes / no
  • CONTRAST_UNVERIFIED items checked by hand, with the lowest ratio found:

4. Not covered by the scripts (check by hand)

  • [ ] Screen reader pass on the main flow (labels make sense, errors are announced)
  • [ ] Copy read aloud once: every button says what happens, every error says what to do
  • [ ] Real data: longest real name, largest real list, slowest real network
  • [ ] Source accessibility scan with a11y-audit (alt text, labels, ARIA, headings)

Decision

  • [ ] Ship: no blockers, every waiver has a WCAG exception or a ticket
  • [ ] Ship with follow-ups (list ticket ids):
  • [ ] Not done: blockers remain

Signed: ______________________ Date: ____________