Reverse Engineering Guide

What it does

  • Upload binaries (optionally with sandbox reports), perform static/dynamic feature analysis, and export JSONL for corpus building.

Binary Analysis UI

Running the backend

  • Dev middleware: available during npm run dev.
  • Standalone: npm run reverse-backend (runs tsx server/start.ts).
  • Ensure Node 18.17+ and required binary analysis deps are present.

Inputs & outputs

  • Accepts single or batch uploads; batch export produces JSONL.
  • Use the Reverse Engineer view to submit files and download results.

Safety notes and precautions

  • Treat untrusted binaries as hazardous: even static analysis can trigger unexpected behaviors if tools aren’t fully sandboxed.
  • Prefer running the app (and analysis tasks) inside a VM or isolated container when handling untrusted samples.
  • Keep yt-dlp/puppeteer up to date if used; they also execute network requests and may parse untrusted media pages.
  • Do not run unknown binaries directly on the host; only use the provided analysis pipeline.