Loading...
Know what your AI built.
AuditBuffet is a library of adversarially-tested audit prompts for vibe-coded projects. Run your free Stack Scan in 30 seconds.
How it works
Step 1
Run the Stack Scan prompt in your AI tool
Copy the Stack Scan prompt from AuditBuffet and paste it into Claude Code, Cursor, or any AI coding tool.
Step 2
Paste the JSON output here
Your AI tool produces a structured JSON telemetry block. Paste it into the AuditBuffet submission form.
Step 3
Get insights and benchmark your project
See your scores by category, understand failing checks, and compare against other projects in the benchmark pool.
5 audits. Zero cost. No account required.
Beyond the Stack Scan, you get five full audits covering security, SEO, accessibility, performance, and mobile responsiveness — each with dozens of checks.
“Why not just ask your AI?”
Two reasons most vibe coders stop rolling their own audits.
Our audits learn from every project
Every submission feeds our benchmark pool. We see which checks produce false positives, which severity weights need recalibrating, and where real vibe-coded projects actually fail. Your chatbot writes checks from training data. Our checks are validated against thousands of real results.
Save your tokens for the fixes
At less than $3/month, building your own audit suite costs more in AI credits than a year of AuditBuffet. Skip the prompt engineering, the edge-case testing, and the version maintenance. Spend those tokens fixing what the audits find instead.
What we audit
One subscription. Every audit. It's a buffet — pick the checks that matter for your stack. See pricing
Security
34 checks covering auth, validation, and secrets exposure
SEO
22 checks for meta tags, structured data, and crawlability
Accessibility
28 checks for ARIA, color contrast, and keyboard navigation
Performance
26 checks for Core Web Vitals, loading, and optimization
Code Quality
30 checks for TypeScript strictness, testing, and patterns
Best Practices
Deployment readiness, error handling, and resilience checks
Think your AI covered everything?
These show up in almost every vibe-coded project.
criticalSecurity
API keys exposed in client bundle
Supabase anon keys, Firebase configs, and third-party secrets shipped to the browser where anyone can extract them.
criticalCode Quality
No error boundary — crashes show a white screen
One unhandled exception takes down the entire app. Users see nothing — no error message, no recovery path.
warningSecurity
API routes with no rate limiting
Every public endpoint is an open door for bots to hammer your database, rack up provider costs, or brute-force auth.
warningAccessibility
Form inputs without labels
Placeholder text disappears on focus and screen readers can't identify the field. 1 in 4 users relies on assistive technology.
infoSEO
Missing Open Graph and meta descriptions
Links shared on Slack, Twitter, or LinkedIn show a blank preview. Google has nothing to display in search results.
These are just 5 of 1,697+ checks across 79 audits.
Works with any AI coding tool.
Paste the audit prompt wherever you already work. The output format is the same regardless of which tool you use.
Claude CodeCursorBoltLovablev0WindsurfCopilotAider
Show off your scores.
Pro subscribers get embeddable badges for their sites, READMEs, and app listings. One line of markdown. Auto-updates with every audit.
Available for overall, pack, and individual audit scores. Included with Pro
Start with your free Stack Scan.
No account required. Runs in your existing AI coding tool in under 30 seconds.