vuln-research
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for system fingerprinting using tools like
uname,lsb_release,apache2 -v, anddpkg, which are standard for identifying software versions on a target system. It also utilizesnmapfor network service identification.\n- [EXTERNAL_DOWNLOADS]: The workflow involves fetching vulnerability data and exploit metadata from established repositories including the National Vulnerability Database (NVD), CISA's Known Exploited Vulnerabilities catalog, and GitHub-based PoC aggregators.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external APIs that could potentially contain untrusted content.\n - Ingestion points: Data is retrieved from NVD, Sploitus, and PoC-in-GitHub APIs via
curlinSKILL.mdandreferences/sources.md.\n - Boundary markers: None present; the agent relies on the structured nature of the JSON returned by these services.\n
- Capability inventory: The skill includes the ability to install packages (
pip), perform network scans (nmap), and interact with the Metasploit framework (msfconsole).\n - Sanitization: None; results are processed using
jqwhich provides basic structure validation but does not filter content for prompt instructions.
Audit Metadata