HarshalAndhale9657/ANVIL ? reverse-engineered prompt
Reverse engineered prompt
Build me a web app called A.E.G.I.S. that can take a GitHub repo, analyze the code for security issues, try to reproduce a vulnerability in a safe sandbox, and only move forward if it can prove the exploit actually worked. If it does, generate a fix, commit it to a branch, and open a pull request automatically.
I want a clean frontend that shows the whole run in real time, including progress, status updates, logs, and whether the process is still scanning, verifying, patching, or done. The backend should handle GitHub login, repo access, the automated workflow, and a reliable safety check so it fails closed if anything looks uncertain.
Make it easy to run locally, and include a one command startup if possible. If you need to look up current docs online for any API details or integrations, go ahead.
Are you gonna build this?
make sure you review the code using coderabbit