better-interface

Orchestrates cross-discipline interface reviews across accessibility, layout, writing, typography, color, and UI polish.

Updated Aug 22, 2026
One-click install
npx skills add https://github.com/falentio/cimi --skill better-interface-falentio
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: better-interface
Source: https://github.com/falentio/cimi/tree/main/.agents/skills/better-interface
Command: npx skills add https://github.com/falentio/cimi --skill better-interface-falentio

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Reviewing a user interface consistently across accessibility, layout, copy, typography, color, and visual polish is hard to do in one pass, and findings often lack evidence, severity ranking, or a clear verdict. This Skill consolidates those domain reviews into a single ranked report with file-and-line evidence. ## Core Features & Use Cases - Cross-domain orchestration: Routes the interface through six domain skills (accessibility, layout, writing, typography, colors, UI) and merges their findings into one ranked table capped at 15 findings. - Evidence-based severity ranking: Applies a shared HIGH/MEDIUM/LOW scale with escalation triggers such as missing accessible names, keyboard-inaccessible controls, and contrast failures. - Structured verdict output: Produces a scope and coverage table, consolidated findings with Before/After fixes, verification results, and a final Block or Approve verdict. - Use Case: Before shipping a new settings page, ask for an interface review to receive a ranked findings table citing exact files and lines, with fixes written in the project's own styling conventions. ## Quick Start Review the checkout flow interface for accessibility, layout, typography, and color issues and give me a ranked findings report.

Frequently Asked Questions about better-interface

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

FAQPage Schema
How do I review a web interface for accessibility and design issues?

Request an interface review naming the screen, flow, or feature to inspect. The review covers accessibility, layout, writing, typography, color, and UI polish, then returns one ranked findings table with file-and-line evidence and a Block or Approve verdict.

What severity levels does an interface review report use?

Findings use three levels: HIGH for issues that block tasks, mislead users, or hide content; MEDIUM for harm to comprehension, efficiency, or consistency; LOW for isolated polish. Escalation triggers like missing accessible names or keyboard-inaccessible controls are always HIGH.

Can I review a pull request or uncommitted changes with this?

Change-scoped reviews of branches, pull requests, or uncommitted work belong to the separate interface-review skill, which is user-invoked. This skill handles screen and flow reviews and receives change scopes handed back from interface-review.

What happens if a domain skill is not available during review?

The missing domain is marked Not reviewed and named in the coverage table, and the review continues with the remaining domains. Its rules are never recreated from memory, and no holistic coverage is claimed.

Does the interface review modify my source code?

Reviews are read-only by default and do not edit source files. If you also ask for implementation, the consolidated report becomes the change scope and relevant verification is re-run afterward.