pr-review-comments
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a template for external data, which introduces a surface for indirect prompt injection.
- Ingestion points: The skill processes pull request findings (title and content) which originate from external code analysis or automated tools.
- Boundary markers: Neither the template nor the instructions provide delimiters or 'ignore' directives to prevent the agent from obeying instructions embedded within the review findings.
- Capability inventory: The skill directs the agent to post or draft inline comments based on the findings (SKILL.md).
- Sanitization: The instructions do not involve any validation, escaping, or sanitization of the interpolated finding content.
Audit Metadata