songshikang0111/awesome-REA-cases

What it solves

Reverse engineering is often slow and hidden behind proprietary tools. This collection makes it easier to find concrete, evidence-backed examples of how AI can assist in reviving old devices, rebuilding applications, preserving games, and analyzing software and hardware behavior. It answers the question: what can AI actually do in reverse engineering today?

How it works

The project is a curated list of 817 cases, updated six times daily. Each case includes a source or author report, evidence links, and clear limits — they are not independently reproduced, so readers can judge reliability. Cases are organized by categories (games, hardware, apps, algorithms) and prioritized by practical value, inspectable implementations, and explicit limits. The collection covers the REA project and other AI-assisted work, with full index and maintenance notes.

Who it’s for

Engineers, hobbyists, researchers, and anyone interested in reverse engineering or AI-assisted development who wants real-world examples, potential starting points, or evidence of what works and what doesn’t. It is also useful for contributors who want to add new cases or verify existing ones.

Highlights

• 817 cases and projects, updated six times daily, covering games, hardware, apps, and algorithms.
• Each case links to public evidence and documents limitations, such as untested claims or single-device setups.
• Includes recent picks like a Linux driver improved via Windows reference behavior, a NIST curve seed investigation, and a Civilization 1991 DOS executable reconstruction.
• Classic cases range from AI-assisted cross-game mods (Minecraft content in GTA V) to decoding a DOS binary and rewriting it in Rust.
• Unofficial, MIT-licensed collection with clear scope: only original summaries and public links, no proprietary software or assets.