better-interface

Orchestrates cross-discipline interface reviews and consolidates findings into one ranked verdict.

Updated Aug 2, 2026
One-click install
npx skills add https://github.com/leonardoacosta/agents --skill better-interface-leonardoacosta
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: better-interface
Source: https://github.com/leonardoacosta/agents/tree/main/skills/better-interface
Command: npx skills add https://github.com/leonardoacosta/agents --skill better-interface-leonardoacosta

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Reviewing a screen or flow across accessibility, layout, writing, typography, color, and visual polish often produces disconnected audits with duplicated or contradictory findings. This Skill routes one interface review through every better-* domain skill and consolidates the evidence into a single prioritized report with a clear verdict. ## Core Features & Use Cases - Holistic orchestration: Loads each owning domain skill (better-accessibility, better-layout, better-writing, better-typography, better-colors, better-ui) in a fixed order and applies their rules without duplicating them. - Scoped review modes: Supports quick mode (primary path, HIGH/MEDIUM findings, cap of 5) and full mode (entire scope including empty, loading, error, and narrow-width states, cap of 15). - Change-scoped reviews: Accepts hand-offs from interface-review with resolved change scopes, adding Introduced/Regression statuses and a separate pre-existing findings section. - Use Case: Ask for a holistic review of a checkout flow; the Skill inspects every domain, cites path/to/file:line evidence, lists considered-but-rejected candidates, and ends with a Block, Needs changes, or Approve verdict. ## Quick Start Ask the agent to run a full better-interface review of the checkout flow and consolidate the findings into one ranked report.

Frequently Asked Questions about better-interface

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a holistic interface review across accessibility, layout, and typography?

Invoke the review with a scope and optional mode, such as a full review of a checkout flow. The Skill loads each better-* domain skill in order, inspects the interface, and consolidates all findings into one ranked table with a final verdict.

What is the difference between quick and full interface review modes?

Quick mode covers the primary path through the scope, reports only HIGH and MEDIUM findings, and caps output at 5 findings. Full mode covers the entire scope including empty, loading, error, and narrow-width states, with a cap of 15 findings.

Can this review uncommitted changes or pull requests?

Change-scoped reviews belong to interface-review, which resolves the diff scope and classifies findings. It then hands the review back here, where findings gain Introduced or Regression statuses and pre-existing issues are listed separately outside the cap and verdict.

What happens if one of the domain skills is not installed?

The missing domain is marked Not reviewed with the unavailable skill named, and the review continues with the remaining domains. The Skill never recreates missing rules from memory or claims holistic coverage it did not perform.

Does the interface review modify my source code?

No, reviews are read-only by default. Source edits happen only when you explicitly request implementation, in which case the consolidated report becomes the change scope and verification is re-run afterward.