LING71671/open-reverselab

Reverse engineering tooling that an AI agent can actually drive, not just assist with. The repo ships 197 articles of structured RE knowledge alongside MCP tooling and an automation chain covering CTF challenges, APK analysis, and PE file inspection. The agent-native framing is the real differentiator here: most RE toolkits are human-operated CLIs; this one is explicitly wired for an LLM to call into. That is a meaningfully different design assumption. The knowledge base alone is worth a browse if you are building security-adjacent AI products and want grounded training material or retrieval context. One very honest reservation the maintainer flags themselves: there is a currently-unpatched bug that causes almost every AI model to jailbreak in this context, and they have chosen not to fix it yet. That is not a minor caveat. Treat this as a research artifact and sandbox it appropriately before running it near anything real. -> Best for: AI engineer or ML researcher building security-tooling agents in a sandboxed environment.