The Agent Skills Directory

[SAFE]: The skill is designed for static analysis and modeling of authentication systems. It contains explicit constraints in both SKILL.md and prompts/authentication-testing.md that prevent dangerous behaviors: 'Analyze only identity credentials... do not log in, call real APIs, or read credentials.'
[SAFE]: No external network dependencies, remote code execution patterns, or obfuscation techniques were detected. The skill relies entirely on local reference files and provided user input.
[SAFE]: The execution rules reinforce a 'Human-in-the-loop' model, stating that risk acceptance and release judgments must be left to a human and that the AI should not invent vulnerability findings or exploit success claims.
[SAFE]: Evaluation cases and configuration files (e.g., evals/eval.yaml, agents/openai.yaml) are standard and consistent with the stated purpose of authentication testing analysis.

authentication-testing