Deep implementation master for Premium Design Critic with 17 targeted Codex skills, code, tests and no MCP. Additive to all existing Perfect_AI versions.
Installs into .claude/skills of the current project.
Are you the author of Ztx4 Premium Critic?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/realjustawall-ztx4-premium-critic)
---
name: ztx4-premium-critic
description: "Deep implementation master for Premium Design Critic with 17 targeted Codex skills, code, tests and no MCP. Additive to all existing Perfect_AI versions."
---
# Premium Design Critic
Review real rendered evidence to detect shallow templates and improve clarity, brand and delight without hallucinating benchmarks.
## Mandatory sequence
1. Inspect screenshots, interaction recordings and source requirements at the same viewport.
2. Grade distinct criteria with evidence and known tradeoffs instead of one vague score.
3. Run anti-cliché review: unnecessary blobs, equally weighted cards, missing content hierarchy and excessive motion.
4. Propose 2–3 targeted, testable design revisions with before/after snapshots.
5. Require functional/a11y regression after applying visual critique.
## Skill routing
Choose only technique files needed for this project from `DEEP-IMPLEMENTATION-v2/references/premium-critic.md`. Never indiscriminately load all 300+ skills into context.
## Runbook
Inputs: target brief, rendered screenshots/recordings, interaction QA and brand evidence.
Implementation: score visual hierarchy, content credibility, distinctiveness, motion meaning, typography, color and mobile; propose ranked fix list backed by screenshots.
Tests: no fabricated evaluation evidence, before/after diff, functional/accessibility regression and independent mobile treatment.
## Acceptance gates
1. Each critique includes screenshot evidence and severity.
2. Rejected clichés explained relative to product brief.
3. Recommended redesign tested for regressions.
## Runnable assets
`extensions/DEEP-IMPLEMENTATION-v2/examples` and `tools` (installed main package). See `RUNBOOK-FA.md`.
## Sources
- [github-awesome](https://github.com/frankxai/awesome-design-agent-skills)
- [github-frontend](https://github.com/hueyexe/frontend-agent-skills)
- [playwright](https://playwright.dev/docs/test-snapshots)
