The Agent Skills Directory

[SAFE]: The skill operates as a high-level methodology and documentation guide. It provides instructional templates for mapping attack surfaces in a structured manner.
[SAFE]: No executable code, shell scripts, or automated tools are included in the skill. All instructions are intended for the AI agent to provide as strategic advice to a human operator.
[SAFE]: The BENCHMARK.md file demonstrates the skill's utility by comparing responses with and without the skill, showing that it helps the agent produce more structured and comprehensive methodology tables.
[SAFE]: References to external tools in BENCHMARK.md (e.g., jwt_tool.py) and specific pentesting techniques are provided as educational examples of common security testing workflows and do not involve automated execution or data exfiltration.

attack-surface