The Agent Skills Directory

[SAFE]: The skill is entirely documentation-based and contains no executable scripts or code, posing no direct execution risk to the environment.
[EXTERNAL_DOWNLOADS]: Lists numerous industry-standard security tools and frameworks from well-known and trusted organizations such as MITRE, Red Canary, and Bishop Fox for red teaming and adversary emulation.
[PROMPT_INJECTION]: Contains a collection of prompt injection and jailbreaking payloads (e.g., 'Ignore previous instructions', 'DAN' personas, and encoding bypasses) strictly as reference examples for evaluating the safety of AI/LLM systems.
[COMMAND_EXECUTION]: Describes technical methods for command and scripting execution (aligned with MITRE ATT&CK T1059) as part of the documented cybersecurity red teaming lifecycle.
[DATA_EXFILTRATION]: Outlines common adversary techniques for data exfiltration to educate users on how to test and improve organizational security defenses.

red-teaming