review
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external, potentially untrusted documents including code diffs, specifications, and implementation plans. This creates a surface where malicious instructions embedded within those artifacts could attempt to manipulate the agent's review logic or subsequent workflow steps.
- Ingestion points: The skill reads file contents and
git diffoutputs during the review process as defined inSKILL.md. - Boundary markers: The mandatory
claim→case tableprovides a structured validation layer that requires the agent to map assertions to specific scenarios, which acts as a logical barrier against simple injection attempts. - Capability inventory: The skill facilitates downstream actions such as committing code or initiating debug sessions via other agent skills.
- Sanitization: There is no explicit text sanitization or filtering defined for the artifact content before it is processed by the LLM.
Audit Metadata