critique

Evaluate web designs for UX problems and generate prioritized P0–P3 feedback reports.

1|Updated Oct 27, 2025
One-click install
npx skills add https://github.com/sandrasocial/sselfie-9g --skill critique-sandrasocial
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: critique
Source: https://github.com/sandrasocial/sselfie-9g/tree/main/.agents/skills/critique
Command: npx skills add https://github.com/sandrasocial/sselfie-9g --skill critique-sandrasocial

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Provide structured, director-level feedback on designs by combining an independent LLM design review with deterministic automated detection so teams can quickly find high-impact usability, accessibility, and AI-generated "slop" issues and know exactly what to fix first.

Core Features & Use Cases

  • Dual independent assessments: Runs an LLM-led design review and a deterministic detector (impeccable) independently to avoid bias and surface different classes of issues.
  • Heuristics & cognitive-load scoring: Produces Nielsen heuristic scores, an 8-point cognitive load assessment, and counts of decision points to quantify UX health.
  • Persona-based and edge-case testing: Walks the product through relevant personas and deliberate stress tests to reveal real-world failure modes.
  • Combined, prioritized report: Synthesizes findings, reconciles false positives, lists P0–P3 issues with why/fix/command mappings, and asks focused follow-up questions.
  • Use Case: Run on a marketing landing page, a settings screen, or a component library to get a prioritized action plan and commands to execute targeted fixes.

Quick Start

Ask critique to review https://example.com/settings and produce a combined LLM and detector-based UX report with Nielsen heuristic scores, cognitive-load results, and prioritized P0–P3 fixes.

Frequently Asked Questions about critique

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

FAQPage Schema
How do I run a UX critique on a live web interface to identify usability issues?▼

To run a UX critique, provide the URL or source markup of your web interface. The tool applies heuristic scoring and cognitive load checks, generating a prioritized P0–P3 action plan to resolve usability issues.

What is heuristic scoring and cognitive load assessment in a design review?▼

Heuristic scoring evaluates designs against Nielsen heuristics, while cognitive load assessment measures mental effort using an 8-point check. Together they quantify UX health and pinpoint decision points increasing user friction.

Can I use automated detection to find AI-generated slop in my component library?▼

Yes, automated detection runs a deterministic scan to identify AI-generated slop in your component library. It operates independently alongside the LLM design review to surface different classes of UX and content issues.

How do I test a settings screen design against specific user personas?▼

To test a settings screen, provide the source markup and specify relevant personas. The tool walks through user journeys and deliberate stress tests to reveal real-world failure modes and edge cases.

Does the design critique require source markup to run deterministic scans?▼

Yes, the design critique requires access to source markup to perform deterministic scans. Optional runtime visualization or CLI tooling can be used to enhance the automated detection process.