nuclei
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides various command-line examples for executing the
nucleisecurity tool. These include scanning single targets, using host lists, and applying specific vulnerability tags.\n- [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection because the agent processes output generated from external, potentially hostile network targets.\n - Ingestion points: The agent is instructed to triage results from output files (e.g.,
contoso-nuclei.txt) that contain data fetched from external web servers.\n - Boundary markers: Absent. No specific delimiters or instructions are provided to help the agent distinguish between tool metadata and potentially malicious instructions embedded in the scanned target's responses.\n
- Capability inventory: The skill enables the execution of CLI tools via shell commands as defined in the
SKILL.mdfile.\n - Sanitization: Absent. There is no mention of filtering or sanitizing the tool output before it is evaluated by the agent.
Audit Metadata