defender
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from pentesting findings or local scan results, which provides a surface for indirect prompt injection attacks.
- Ingestion points: The skill ingests data via the
$ARGUMENTSvariable and by reading security scan results from the project directory as described inSKILL.md. - Boundary markers: There are no specific delimiters or instructions provided to the agent to help it distinguish between its primary directives and the potentially malicious content within the findings it analyzes.
- Capability inventory: The skill triggers a subagent (
agent: defender) to analyze and interpret the provided data, involving reasoning over external inputs. - Sanitization: No sanitization, filtering, or validation mechanisms are present to clean the input findings before they are processed by the subagent.
Audit Metadata