The Agent Skills Directory

[SAFE]: The skill executes a local Python script (scripts/check_offer_terms.py) that uses only standard libraries. No external network requests, remote code execution, or unauthorized command patterns were found.
[DATA_EXFILTRATION]: Although the skill processes sensitive personal information (PII) such as wages and employment contracts, it lacks network access to exfiltrate data. The instructions explicitly direct the agent to minimize PII usage and avoid saving unnecessary data.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data through user-provided job offer details. This ingestion surface is considered safe as the processing script validates input types and performs static comparisons without the use of dangerous dynamic execution functions like eval() or exec().

offer-terms-check