The Agent Skills Directory

[SAFE]: The automated scanner alerts regarding 'loldrivers.io' and the malware classification of 'kernel-driver-reverse.md' were reviewed. 'loldrivers.io' (Living Off Trusted Drivers) is a well-known community project for tracking vulnerable drivers, and the associated markdown file contains educational documentation on driver reversing and rootkit analysis. These are classic false positives for security documentation.
[SAFE]: Tool installation instructions in 'SKILL.md' (e.g., 'frida-tools', 'angr', 'z3-solver', and 'pycdc') refer to standard, industry-recognized security tools used for reverse engineering. Repositories cited, such as 'zrax/pycdc' and 'mandiant/GoReSym', are established open-source projects.
[SAFE]: The instructions provided for the AI agent (e.g., in 'references/re-agent-workflow.md') are focused on structured problem-solving phases (triage, static, dynamic, synthesis). There are no attempts to bypass safety filters or ignore initial instructions.
[SAFE]: Examples of sensitive file access (e.g., reading '/proc/self/status' or '/proc/self/maps') and shell commands are presented strictly as educational demonstrations for identifying and bypassing anti-debugging techniques in a sandbox environment.
[SAFE]: No obfuscation, hidden URLs, or malicious persistence mechanisms were found in any of the 25 files analyzed.

reverse-engineering