meta-reviewing-infra-reviewing
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data in the form of code diffs provided by users or external systems. This creates an inherent attack surface where a malicious diff could contain hidden instructions intended to subvert the agent's review logic.
- Ingestion points: Untrusted content enters the agent's context whenever it is asked to review a diff touching CI/CD workflows, Dockerfiles, or IaC configurations (SKILL.md).
- Boundary markers: The instructions do not specify the use of delimiters or specific 'ignore' markers to encapsulate the untrusted diff content during analysis.
- Capability inventory: The skill itself is instructional, but the agent's environment likely includes file-system access or tool execution capabilities that could be targeted by an injection.
- Sanitization: There is no evidence of automated sanitization, filtering, or validation of the input diff content before the agent processes it for review.
Audit Metadata