Works with the AI tools you already use
Claude Code
Cursor
Codex CLI
GitHub Copilot
Gemini CLI+20 moreClaude CodeCursorCodex CLIGitHub CopilotGemini CLI
OpenClaw
Manus
Grok+17 agentsUX-UI Audit
Turn UI screenshots and flows into prioritized, evidence-backed UX audits with actionable engineering requirements.
$5.99
/moSubscription · cancel anytime
Secure checkout via Stripe
- Always the latest version
- Delivered live to your agent via MCP
- Cancel anytime, access ends at period end
UX-UI Audit
Example session with this skill installed
Audit this checkout screen screenshot. Users are complaining that they aren't sure if their order went through. The artifact is a mobile web view of the 'Place Order' page.
- Read your context and instructions
- Compiled the ux-ui audit
- Generated the UI component
I have audited the 'Place Order' screen for the mobile web view. The audit focuses on feedback loops and state transitions.
Audit Frame
- Mode: Screen Audit
- Artifact: Static Screenshot (Mobile Web)
- Confidence: Medium (Static view prevents inspection of network latency behavior)
Executive Diagnosis
- The primary friction is a lack of visual affordance for the submission state and an ambiguous final action.
- Information hierarchy favors secondary order details over the primary transaction control.
Prioritized Findings
S1 | High confidence | Checkout / Submission
Users risk duplicate orders due to lack of submission feedback.
Evidence: The 'Place Order' button remains visually active after tap with no loading spinner or progress overlay.
Impact: Users may tap the button multiple times, leading to duplicate charges and backend errors.
Root cause: The interface does not lock the UI thread or provide an intermediate 'Processing' state.
Recommendation: Disable the button immediately upon tap and replace the label with a loading indicator.
Acceptance criteria: Button is programmatically disabled until a response is received; a screen reader announces the 'Processing' status.
S2 | Medium confidence | Layout / Hierarchy
Primary CTA is occluded by the virtual keyboard on smaller viewports.
Evidence: The 'Place Order' button is positioned at the bottom of a long scrollable form without a sticky footer.
Impact: Users must scroll past redundant order summaries to find the final action.
Root cause: Viewport-aware layout constraints are missing for the primary action.
Recommendation: Implement a sticky footer for the primary action or move the button above the fold.
Acceptance criteria: The primary CTA is visible on a standard 390x844 viewport without scrolling.
Next steps
- Verify if the backend supports idempotent requests to prevent duplicate processing.
- Test the button state with a slow 3G network simulation.
- Implement a
submission-lockedstate in the checkout component.
ux-ui-audit.tsx
TSX · React component
Example file from a real run - the skill writes it into your workspace.
Subscription · always the latest version
- Secure checkout via Stripe
- Cancel anytime
- Delivered live via MCP (optional)
What you get
About this skill
UX audits often devolve into subjective aesthetic critiques or vague lists of "bad UI" that developers can't act on. This skill transforms product reviews into diagnostic, evidence-backed engineering requirements. It eliminates guesswork by mapping every interface flaw to a specific user consequence and a technical root cause.
What it does
- Evidence-based auditing identifies issues by tracing exact sequences from user action to observed response and consequence.
- Severity calibration categorizes findings from S0 (Critical/Blocking) to S3 (Minor/Polish) based on task impact and frequency.
- Root cause analysis isolates the underlying system or design decision rather than just reporting visual symptoms.
- Actionable recommendations provides minimal sufficient changes including behavior, content, and clear acceptance criteria.
- Journey mapping visualizes the critical path to identify where user mental models diverge from the product logic.
How it works
- Frame: Define the product, user role, and audit mode (Screen, Journey, System, or Release).
- Inspect: Provide screenshots, URLs, or source files for the agent to map states and transitions.
- Diagnose: The agent evaluates the artifact against usability lenses like navigation, state completeness, and accessibility.
- Report: Receive a prioritized list of findings formatted for immediate entry into a backlog or PR.
Frameworks & tools
This skill utilizes WCAG 2.2 standards, cognitive load principles, and heuristic evaluation frameworks. It is designed for web, mobile, and AI-integrated applications.
Why this beats prompting it yourself
Generic prompts yield "make it pop" advice or generic lists of UX laws. This skill enforces a rigorous evidence register, ensuring every critique is grounded in observed behavior and technical feasibility rather than subjective opinion.
Use cases
- Reviewing a checkout flow to identify why users abandon at the payment step.
- Auditing a complex B2B dashboard for information density and wayfinding issues.
- Checking a new feature release for accessibility barriers and failure state handling.
- Evaluating AI-generated interfaces for provenance, control, and recovery paths.
Known limitations
Cannot confirm WCAG compliance or legal standing from static screenshots alone. Requires inspectable artifacts; it will not fabricate an audit from a text description of a product.
How to install
Works the same in every agent - Claude, Cursor, Codex, Copilot and 20+ more.
- 1
Download the ZIP
Free skills download straight away. Paid skills unlock right after purchase.
- 2
Unzip into your skills folder
Every agent reads skills from one folder on your machine. Drop the unzipped folder in there.
- 3
Ask your agent to use it
Restart the agent if it was already running. It picks the skill up automatically - no config needed.
Skills folder by agent
Click the path to copy it. Create the folder if it does not exist yet.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scanned
Verified clean 2 days ago