audit-code-security
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data by auditing external code, diffs, and pull requests, creating a surface for indirect prompt injection.
- Ingestion points: The agent ingests code, repository metadata, and pull request content from the target environment as defined in the scope section of
SKILL.md. - Boundary markers: The instructions do not specify the use of delimiters or specific boundary markers to isolate untrusted code from the agent's internal reasoning.
- Capability inventory: The skill is explicitly limited to read-only actions. It forbids installing tools, running destructive tests, probing external network targets, or modifying repository configurations. It restricts the agent to using tools already available in the environment.
- Sanitization: There are no instructions for sanitizing or filtering ingested code before analysis, though the read-only constraint and focus on evidence-based verification mitigate the risk of accidental execution.
Audit Metadata