The Agent Skills Directory

[SAFE]: The skill serves as a standard operating procedure (SOP) for cybersecurity analysts. It contains no executable scripts, shell commands, or tool configurations that could be used maliciously.
[PROMPT_INJECTION]: The instructions do not contain any patterns typical of prompt injection, such as attempts to override system prompts or bypass safety filters.
[EXTERNAL_DOWNLOADS]: The skill references several well-known cybersecurity resources (MITRE ATT&CK, VirusTotal, Hybrid-Analysis, abuse.ch) and official GitHub repositories for security tools (CAPE, Yara-Rules). These references are informational and do not involve automated script execution or untrusted downloads.
[DATA_EXFILTRATION]: There are no network operations or commands designed to exfiltrate sensitive data. Mention of sensitive file paths (like startup folders or registry keys) is strictly in the context of detecting malware persistence, not for unauthorized access by the skill itself.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns or dynamic code evaluation methods were found.

malware-triage