impeccable vs taste-skill
Design vocabulary skill for coding agents: commands like polish, bolder, critique and audit, plus a 62-rule deterministic detector that catches AI-slop UI before it ships. — versus — Agent skill that art-directs coding agents into polished company websites — one design system, real photos and icons, restrained motion — verified in a real browser by a finish gate.
Both stop agents shipping generic AI-looking pages; taste-skill art-directs whole sites from a brief with a browser finish gate, Impeccable gives targeted design commands and audits for existing UI.
| impeccable | taste-skill | |
|---|---|---|
| Stars | 113k | 119k |
| Forks | 6.8k | 8.0k |
| Language | Rust | JavaScript |
| License | Apache-2.0 | MIT |
| Last activity | today | 11 days ago |
| Topics | skills | skills |
| Curated connections | 3 | 4 |
impeccable — the curator's take
Best when iterating on UI that already exists: name a move and a target, get a focused playbook, then critique/audit with ranked findings — the slop detector needs no LLM. Works across ~20 agents. It steers rather than art-directs from scratch; for whole marketing sites built from one brief with browser verification, taste-skill goes further.
taste-skill — the curator's take
Install when your agent ships landing pages that all look like the same purple-gradient template: it enforces real art direction and checks the result in Chrome instead of trusting the agent's word. Opinionated by design — it imposes its own taste, so skip it when you already have a brand system to follow (feed that in instead) or when you're building app UI/dashboards rather than marketing sites. Needs Node 22+ and a local Chromium.