CODE CLINIC
Check your app before launch.
Your code stays with you.
Your own AI agent runs 14 checks inside your project. You paste its short report here, and this page tells you what to fix first and how.
Copy the prompt
It tells your agent what to look at and how to write the report.
Run it in your project
Paste it into Cursor, Claude Code or Codex. The agent reads the project and changes nothing.
Paste the report here
The page reads it in your browser and shows what to do next.
01 / PROMPT
One prompt, 14 checks.
The checks cover secrets, access to data, how the app is built and started, and how safe it is to change.
The prompt asks for a report without passwords, keys or personal data. Only statuses and short notes leave your project.
Which agent to use
The check needs an agent that can read the files of your project.
- Tools
- Cursor, Claude Code, Codex
- Best results
- Claude Opus 5.5, GPT-6.1 Sol, Gemini 4 Carbon
- Also fine
- Claude Sonnet 5.5 and other mid-tier models. Expect more UNKNOWN answers.
Built the app in Lovable, Bolt or a similar builder? Paste the prompt into its chat. If it cannot see every file, export the project to GitHub and open it in one of the tools above.
Model names last checked in November 2026.
02 / REPORT
Paste the report from your agent.
Paste the whole answer. The page finds the report in it.
The report is read in your browser. It is not sent or stored anywhere, unless you choose to send it to us below.
Want the full handover instead?
The handover prompt on the home page produces a longer technical summary of your project for a hosting proposal.
