Principal UX Designer
Usability, accessibility, hierarchy, and visual polish — mapped to WCAG.
Your instant principal UX expert
Your web app, reviewed like a flagship product.
Three expert reviews (design, engineering, and growth) in one prioritized report. It runs instantly in your browser, so your code never leaves your device. Go deeper with Claude on your own API key, and tune it to your audience with custom rules and uploadable expertise. A generic checker gives you one score. This gives you a verdict.
Built for indie hackers, agencies, and product teams
See the full diagnosis free — just your email to start. Pro unlocks the code-native fix for every finding.
Private by design — local audits never leave your browserFree diagnosisOpen engine you can inspect
Full-Stack Engineer
Performance, security, correctness, and tech debt — with fixed snippets.
Growth Strategist
Onboarding friction, trust signals, retention loops, and A/B hypotheses.
Why it's different
Not a generic checker. A reviewer you control and tune.
Private local engine
Run a full three-lens audit instantly, right in your browser. Your code never leaves your device and no API key is required — most tools upload everything to a server first.
Deeper review, your own key
Want more? Add your own Anthropic API key for a deeper Claude review that reads code and screenshots — sent browser-to-Anthropic with no middleman. You own the data and the cost.
Tuned to your world
Set your audience, switch on custom rules like AAA contrast, and upload expertise skills that teach the reviewer your domain. The audit re-weights around your real users — not a generic checklist.
How it works
Three steps to a shippable roadmap
01
Add your product
Drop in HTML, CSS, and JavaScript files, paste code directly, add screenshots for a visual review, or point UXexpert at a live URL or GitHub repository. Add design systems or brand guidelines to tune enforcement.
02
Tune the review
Choose the audience you are building for, set custom rules, and load expertise skills that teach the reviewer your domain — from telco BSS collections consoles to patient portals. The audit re-weights itself around your users.
03
Ship the fixes
Every finding is scored by severity, effort, and impact, and arrives with a concrete fix — often a code snippet you can paste in as-is. Export the report and get to work.
Live playground
Run an audit right now
Paste a URL, drop files, or paste code — then hit Run. The local engine is instant and free; add your Anthropic API key for a deeper Claude review. New here? Load the sample app to see a full report in one click.
01
Add your product
The page is fetched for you — public GitHub repos work too (a branch or folder link is fine).
A fine-grained token with repository read access unlocks private repos and a 5,000-requests-per-hour limit. Sent only to GitHub, and kept only for this browser session — it's cleared when you close the tab.
or drop files / screenshots
Drop files here or click to browse
.html · .css · .js · screenshots (.png, .jpg) — multiple files welcome
or paste code
02
Tune the review
optionalTeach the reviewer your domain: who your users are, what to research and validate, and what good design means in your world. Skills are applied in depth by the Claude AI engine.
03
Choose an engine
Used only for direct calls from this page to api.anthropic.com. If remembered, it is stored in this browser's local storage on this device — never on a server. That's convenient but means any script running on this page could read it, so we recommend a key with a spending limit and leaving "remember" off on shared computers. Create a key at console.anthropic.com.
04
Get your report
Nothing to audit yet — add your product in step 01.
Nothing to paste? Load the sample app
Something went wrong
Your report will appear here
Findings from all three reviewers, scored and prioritized,with fixes you can apply today.
Features
Depth that generic checkers can't reach
True multi-role synthesis
Not three separate reports — one prioritized roadmap where design, engineering, and growth trade-offs are weighed against each other.
Code-native fixes
Every technical finding ships with a corrected snippet you can paste straight into your codebase.
Audience tuning
Auditing for Gen Z mobile users in emerging markets is a different job than auditing for enterprise buyers. Priorities re-weight to who you serve.
Expertise skills & custom rules
Upload skills that make the reviewer an expert in your domain and your users, plus rules like AAA contrast and minimal motion. The audit enforces them.
Growth playbook
A retention roadmap, A/B hypotheses, and conversion fixes grounded in what is actually on your page.
One-click exports
Markdown export today. PDF, Jira tickets, Figma comments, Notion pages, and shareable report links are on the roadmap.
Version comparison
Re-run after each iteration and compare scores. Automatic history tracking is on the roadmap.
Private by design
Local analysis never leaves your browser. AI reviews go directly from your browser to Anthropic with your own key — no middleman server.
API for teams
On the roadmap: wire audits into CI so every pull request gets a design, code, and growth review before merge.
Pricing
The diagnosis is free. The fixes are Pro.
Run a private, three-lens review free — in your browser, tuned to your audience, as deep as you want with your own Claude key. Pro unlocks the code-native fix for every finding at $49/month or $490/year (two months free). Cancel anytime.