pentest-expert
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill primarily consists of documentation and reference materials for security professionals. It does not contain executable code or instructions that bypass safety protocols.
- [COMMAND_EXECUTION]: The document includes numerous examples of shell commands for reconnaissance and scanning tools like nmap, gobuster, and nuclei. These are presented for educational and professional use within the context of security assessments and are not executed automatically.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted targets (URLs, domains, and IP addresses) which could potentially contain malicious payloads. However, the instructions are static and do not indicate automated processing of results that could lead to agent compromise, resulting in a low-risk profile.
- Ingestion points: Target URLs and domains provided by the user for scanning (e.g.,
target.com). - Boundary markers: Not explicitly defined in the methodology.
- Capability inventory: Uses network scanning and enumeration tools via shell commands.
- Sanitization: Not applicable as the skill provides templates for human-driven tool execution.
Audit Metadata