technical-quality-perspective
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary purpose is to ingest and process untrusted external data, such as code diffs, architecture documentation, and test reports, which creates a surface for indirect prompt injection if those materials contain malicious instructions.
- Ingestion points: Untrusted data enters the agent's context through the review stages defined in SKILL.md and the specific instruction sets in the prompts/ directory.
- Boundary markers: The skill provides a rigid 'Report contract' requiring specific sections (Summary, Facts, Evidence, etc.) and explicitly commands the agent to distinguish supplied facts from its own inferences, which acts as a deterrent to model deviation.
- Capability inventory: The skill does not define any tools, shell access, or network-capable commands, significantly limiting the potential impact of an injection.
- Sanitization: No technical sanitization or input filtering is implemented for the analyzed text, relying entirely on the agent's internal safety guidelines and the provided prompts.
Audit Metadata