The Agent Skills Directory

[SAFE]: The skill provides documentation, detection queries, and rule templates for defensive security operations. All code snippets (SQL, SPL, KQL, YARA, Sigma) are standard for security monitoring and do not contain malicious payloads or exfiltration logic.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze external data such as log files and threat intelligence feeds. While these sources are inherently untrusted and could contain adversarial payloads (e.g., log entries designed to manipulate the LLM), the skill itself acts as a static knowledge base and does not implement automated tool execution or dangerous file system/network operations that would facilitate an exploit. Standard security boundaries apply.

threat-hunting