skills/mukul975/anthropic-cybersecurity-skills/analyzing-powershell-script-block-logging/Gen Agent Trust Hub
analyzing-powershell-script-block-logging
Pass
Audited by Gen Agent Trust Hub on May 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: No security issues detected. The skill is designed for defensive security operations and forensic investigation.
- [EXTERNAL_DOWNLOADS]: The skill utilizes well-known, legitimate Python libraries for parsing Windows event logs and XML data.
- Evidence:
pip install python-evtx lxmlinSKILL.mdand imports inscripts/agent.py. - [PROMPT_INJECTION]: While the skill processes untrusted data from forensic log files, it does so in a parser-only context with no execution capabilities, posing no risk to the agent environment.
- Ingestion points:
scripts/agent.pyreads data from.evtxfiles viaparse_evtx_4104. - Boundary markers: None explicitly used for processed content.
- Capability inventory: File system read/write only; no network access or command execution.
- Sanitization: Content is extracted and analyzed via regex/entropy; no specific sanitization is applied to the raw script text before JSON output.
Audit Metadata