skills/mukul975/anthropic-cybersecurity-skills/analyzing-windows-shellbag-artifacts/Gen Agent Trust Hub
analyzing-windows-shellbag-artifacts
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements standard digital forensics procedures for interpreting Windows Shellbag registry keys. The provided Python scripts perform local binary parsing and data transformation without suspicious network activity.
- [EXTERNAL_DOWNLOADS]: References to external tools (SBECmd, ShellBags Explorer) and libraries (regipy) target reputable community sources and official repositories. These are documented neutrally without alarming language.
- [PROMPT_INJECTION]: Indirect prompt injection surface detected. The skill processes external registry hives and CSV files without explicit boundary markers. This is a low-risk inherent characteristic of forensic data processing tools. Evidence: 1. Ingestion: agent.py (target hive path), process.py (CSV input). 2. Boundary markers: Absent. 3. Capabilities: File read/write. 4. Sanitization: ASCII decoding with error replacement in agent.py.
Audit Metadata