Hi Product Hunt 👋 I build apps with AI coding tools every day. They're incredible — and they also hardcode API keys into client bundles, commit .env files, and leave source maps on production domains. I know because I kept doing it myself, and because the first thing I built to check for it kept finding the same five mistakes in almost every vibe-coded app I scanned. So I turned it into OpzyAI: 🔍 Free Vibe Check — paste your live URL, get a 0–100 Launch Readiness score in ~15 seconds. No account, no install. It's passive: it only looks at what your site already shows every visitor — leaked keys in the JS bundle (OpenAI, Anthropic, Stripe, Supabase), exposed .env / .git / source maps, missing security headers. 🛠️ Fixes, not jargon — every finding is plain English: what's exposed, why it matters, and a fix you can paste straight into Cursor or Claude Code. 🏅 Badge — score 80+ and you get an embeddable "Vibe Check ✓" badge for your README. 🤖 MCP — run it from inside your editor. The free local server (npx -y @opzyai/mcp) checks your working tree for secrets before you push. The Pro server deep-scans your repo (dependency CVEs, SAST, git-history secrets) and hands your agent the exact fix to apply — a precise package.json upgrade or a key-rotation runbook. OpzyAI never writes to your repo; your editor applies, you review. Honest scope note: a passive scan can't see server-side bugs — a clean score means "not leaking the obvious stuff," not "secure." That's what the deep repo scan is for (Pro, €29/mo, 14-day trial). Scan your app and tell me what it gets wrong — false positives are bugs and I'll fix them same-day: https://www.opzyai.com/scan?utm_...

OpzyAI

Finds what your AI-built app leaks — your editor fixes it

Finds what your AI-built app leaks — your editor fixes it