UX & Accessibility Website Audit Tool
An AI-powered analyzer built with Lovable
Paste any website URL and instantly get a structured audit covering UX and accessibility issues - with scores, prioritized problems, and actionable recommendations, all in one clear interface.
No-code/AI-assisted development
How it works?
Users paste any public website URL and the tool analyzes it across UX, accessibility, and performance dimensions, returning a structured report in a clear card-based interface with visual severity cues.
Results include:
* UX Score - overall usability rating covering layout, interactions, and design patterns
* Navigation & Readability scores - how easily users find their way and consume content
* Accessibility (WCAG 2.1 AA) - compliance check with specific violations flagged by severity
* Performance issues - including real metrics like Largest Contentful Paint
* Conversion optimization tips - actionable suggestions to improve user actions
* Prioritized issues - categorized as Critical, Warning, and Informational
How it works?
Users paste any public website URL and the tool analyzes it across UX, accessibility, and performance dimensions, returning a structured report in a clear card-based interface with visual severity cues.
Results include:
* UX Score - overall usability rating covering layout, interactions, and design patterns
* Navigation & Readability scores - how easily users find their way and consume content
* Accessibility (WCAG 2.1 AA) - compliance check with specific violations flagged by severity
* Performance issues - including real metrics like Largest Contentful Paint
* Conversion optimization tips - actionable suggestions to improve user actions
* Prioritized issues - categorized as Critical, Warning, and Informational
How I built it
Built entirely through prompting in Lovable - no code written. The process involved multiple iterations, refining both the AI logic and the interface design through conversational prompts. I focused on defining the scoring rubric, issue categories, and output structure to ensure results are consistent and immediately actionable. As the tool evolved, I prompted integrations with Google Lighthouse and axe-core to enable real performance and accessibility data - moving beyond AI inference toward measurable, standards-based analysis.
Note:
This is a prototype currently in progress. The tool already performs real HTML and metadata analysis and returns WCAG 2.1 AA compliance checks, performance metrics, and specific accessibility violations. Lighthouse and axe-core integrations were implemented through prompting but require further testing to validate reliability across different website types. A production-ready version would build on this foundation to deliver fully deterministic, standards-compliant reports.
