pentest-cve-vulnerability-research-helper
Pass
Audited by Gen Agent Trust Hub on May 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides shell command templates for tools like
vulnxandcurlthat interpolate user-supplied data (such as product names, versions, or URLs) into double-quoted strings. This pattern creates a potential command injection surface if the agent fails to sanitize inputs before execution, as shell substitution (e.g.,$(...)) is still possible within double quotes.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it fetches and processes untrusted data from external sources.\n - Ingestion points: Web content fetched via
curland vulnerability intelligence data retrieved viavulnxfrom the ProjectDiscovery API.\n - Boundary markers: The instructions do not define boundary markers or provide 'ignore embedded instructions' warnings for the external content being processed.\n
- Capability inventory: The skill utilizes shell execution for
vulnx,curl,jq, andgrep, and it manages a third-party API key (PDCP_API_KEY) via environment variables.\n - Sanitization: While the skill uses
jqandgrepto filter or parse data, it lacks explicit sanitization of the fetched content to prevent it from influencing agent behavior.
Audit Metadata