review-loop
Pass
Audited by Gen Agent Trust Hub on May 3, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes project plans and code artifacts, which creates a surface for indirect prompt injection.
- Ingestion points: Content from project plans or implementation work (SKILL.md).
- Boundary markers: No explicit boundary markers are defined to isolate or ignore embedded instructions within the artifacts being reviewed.
- Capability inventory: The agent is instructed to modify files to fix findings, commit changes to git, and use the GitHub CLI to create issues (SKILL.md).
- Sanitization: No sanitization or validation of the ingested artifact content is specified before processing or executing fixes.
- [COMMAND_EXECUTION]: Utilizes the GitHub CLI (gh) to record review findings as issues. This operation uses a well-known service and is consistent with the skill's stated purpose.
Audit Metadata