how
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external codebase content which inherently creates an indirect prompt injection surface if source files contain malicious instructions disguised as comments.
- Ingestion points: The skill dispatches explorer subagents to read and analyze source code as defined in
SKILL.mdandreferences/explorer-brief.md. - Boundary markers: The prompts do not specify the use of clear delimiters or instructions to ignore potential instructions embedded within code comments or documentation.
- Capability inventory: The skill is focused on analysis and documentation synthesis; it does not include instructions for sensitive data exfiltration or arbitrary command execution.
- Sanitization: The
references/explorer-brief.mdmandates that findings must be traced to specific file paths, function names, and line numbers, providing a factual grounding that helps mitigate the impact of adversarial content within the code.
Audit Metadata