The Agent Skills Directory

[SAFE]: Extensive security analysis confirms the skill contains no malicious instructions, prompt injections, or persistence mechanisms. The skill follows best practices for sensitive HR-related tasks.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it ingests untrusted data from both user inputs (role descriptions) and external public sources (BLS, Census). However, the risk is mitigated by anchoring the analysis in the Standard Occupational Classification (SOC) system and using rigid output templates that prevent external content from influencing the agent's core instructions.
[EXTERNAL_DOWNLOADS]: The skill is configured to query authoritative and trusted government repositories, including the Bureau of Labor Statistics (BLS), the Office of Personnel Management (OPM), and the US Census Bureau. These references are essential to the skill's functionality and do not represent a security threat.

comp-benchmark