pentester
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill passes user-supplied input from the $ARGUMENTS variable directly into the agent's context, creating a potential attack surface for malicious payloads embedded in targets.
- Ingestion points: The skill uses the $ARGUMENTS placeholder in SKILL.md to ingest untrusted data defining the pentest target or task.
- Boundary markers: The instructions lack explicit delimiters or warnings to the agent to treat the target input as data rather than instructions.
- Capability inventory: The description indicates that the subagent is capable of "multi-tool attack workflow orchestration," implying it has access to powerful tools that could be misused if the agent is compromised via injection.
- Sanitization: No sanitization, validation, or escaping logic is defined to check the target input for embedded instructions or malicious patterns.
Audit Metadata