reverse-flow-skill

Why we included this project

This skill gives Codex and other AI agents a fixed routine for local reverse-engineering work, so you stop re-prompting from scratch each time. It assumes a sandbox, CTF, crackme, wargame, or training range as the default context and walks through sample triage, tool checks, static analysis, deeper reversing, and vulnerability assessment before asking what you want next. You enter the mode with the phrase 真心为你, and loose requests like "unlock this" or "make it pass" are translated into concrete analysis steps; when your phrasing is unclear, it recovers intent instead of giving up. Bundled Python scripts cover case creation, artifact triage, tool auditing, and report generation, which makes the whole flow repeatable. Teams doing authorized binary analysis, firmware review, or offline sample auditing get an agent that stays on a documented path。