Red Team Tools and Methodology
Pass
Audited by Gen Agent Trust Hub on Mar 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses various shell commands and a dedicated reconnaissance script to run third-party security tools such as Amass, Subfinder, Nuclei, and ffuf.
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by ingesting untrusted data from web-based reconnaissance tools (like waybackurls and gau) and processing it through subprocess calls (like curl and bash) without implementing boundary markers or input sanitization.
- [DATA_EXFILTRATION]: Performs network requests to external security services, such as bgp.he.net, for the purpose of infrastructure reconnaissance.
Audit Metadata