Broken App Diagnosis (Lovable, Bolt, v0, Replit)
Pricing
$5.00 / complete app diagnosis report
Broken App Diagnosis (Lovable, Bolt, v0, Replit)
GitHub repo URL in, plain-English diagnosis out. Clones the app, installs and builds it in a clean environment, renders it in a headless browser, and reports exactly what is broken and why: missing env vars, dependency conflicts, build blockers, white screens, leaked keys, SPA routing gaps.
Pricing
$5.00 / complete app diagnosis report
Rating
0.0
(0)
Developer
Dylan Caponi
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
6 days ago
Last modified
Categories
Share
Broken App Diagnosis for AI-Built Apps
Your Lovable, Bolt, v0, or Replit app worked in the tool, then broke the moment it left: blank screen after deploy, build errors, "works for me but not for anyone else". This Actor tells you exactly what is wrong, in plain English, in a few minutes.
Give it a public GitHub repo URL. It will:
- Clone your app into a clean, isolated environment (nothing from your machine, no stored logins, no env vars).
- Check the code statically: missing
.env.example, env vars your code needs but never documents, config modules that crash before React mounts, possible secret keys committed to client code, missing SPA rewrite config that 404s your routes after deploy. - Install dependencies with lifecycle scripts disabled, and classify failures (the classic
ERESOLVEconflict AI tools create gets a specific explanation). - Build the production bundle and pinpoint blockers like
Cannot find module, which usually means the AI used a package it never declared. - Actually run your app with a stack-aware runtime check, catching the "builds fine, ships a white screen" failure class:
- Vite/SPA builds render in a headless browser and we measure whether your UI actually mounts.
- Next.js apps boot with
next startand get fetched over HTTP, so server crashes on startup, 500s, and "Application error" pages are caught. - Node/Express servers (and fastify, koa, hono) boot with
npm start; we detect crash-on-boot, servers that never start listening, and error responses.
Why a clean environment matters
Most "it broke after export" problems are invisible on the machine that grew the app, because that machine has the env vars, the logins, and the node_modules that make it work. This Actor reproduces what a new user, a fresh deploy, or a collaborator actually experiences. That is where your bug lives.
Input
| Field | Type | Description |
|---|---|---|
repoUrl | string | Public GitHub repository, e.g. https://github.com/owner/repo |
Output
- A ranked, plain-English markdown report in the run's key-value store (
OUTPUT), worst problems first, each with what it means and how it gets fixed. - A structured summary row in the dataset: blocker, risk, and note counts plus the verdict, so you can wire this into automations.
What it covers
- Lovable, Bolt.new, and v0 exports (Vite + React + Supabase stacks get extra checks)
- Replit apps pushed to GitHub, including the fullstack template (Vite client + Express server)
- Any Vite, Next.js, or Node/Express JavaScript/TypeScript project with a standard install/build/start flow
Repos that are private, deleted, or not JavaScript projects fail fast with a clear message.
Who built this
Built by Vibe Code Rescue, a service that fixes broken AI-built apps. The same engine powers our free diagnosis intake; if the report finds something you want fixed rather than explained, that is what we do.
