reverse-engineering

SUSPICIOUS: the skill is purpose-aligned as a reverse-engineering playbook, but it gives an AI agent high-risk offensive analysis capability: anti-debug bypass, memory dumping, binary patching, key interception, and malware-oriented workflows. Install sources are mostly legitimate developer channels, so this is not confirmed malware, but it is a high-risk security/exploit skill whose capability exceeds ordinary assistant scope and should not be treated as benign.

The skill is internally consistent with its stated reverse-engineering purpose and does not show credential theft, covert exfiltration, remote installers, or third-party proxy routing. However, it equips an AI agent with offensive reverse-engineering and anti-detection browser automation techniques for risk-control/captcha systems, making it a high-risk security workflow despite low evidence of malware.

The provided fragment is high-risk dual-use offensive content. It contains a concrete Frida-based jailbreak-detection bypass that intercepts `access()` and forces failure (`retval.replace(-1)`) for known jailbreak-related paths, plus extensive step-by-step guidance for dynamic instrumentation/injection concepts, firmware extraction, kernel/driver/eBPF reversing, and automotive RE. The fragment does not, by itself, demonstrate credential theft, persistence, or data exfiltration, but it clearly enables evasion and follow-on attack development. Treat as potentially malicious/attack-enabling unless verified as non-executed documentation within a benign repository.