The Agent Skills Directory

[DATA_EXFILTRATION]: The skill implements strict privacy controls in the 'Individual Information and Authority Boundaries' section. It explicitly prohibits sending a user's sensitive history or personal identifiers to external search engines, job platforms, or web forms, requiring all research to be conducted with generalized parameters to mitigate data leakage risks.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted data from user conversations and external job descriptions (Ingestion points). It provides specific instructions to separate stated facts from generated hypotheses (Boundary markers) and restricts capabilities to conversational analysis and manual source verification (Capability inventory). It further requires the sanitization of search queries to protect user identity (Sanitization).\n- [SAFE]: No malicious code, dynamic execution, persistence mechanisms, or obfuscation techniques were identified within the skill files. The skill relies entirely on markdown-based instructions for conversational guidance, posing no technical risk to the host environment.

career-options