pr-comments
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from pull request comments, which creates a surface where instructions embedded in those comments could potentially influence agent behavior.
- Ingestion points: The instructions in
SKILL.mdspecify fetching inline threads, review summaries, and general comments. - Boundary markers: There are no explicit instructions for the agent to use delimiters or ignore instructions found within the content of the comments.
- Capability inventory: The skill is empowered to implement and validate code changes (Step 4), which involves file modification and likely the execution of testing tools.
- Sanitization: The instructions do not define any sanitization, filtering, or validation steps for the incoming comment data.
Audit Metadata