plan-reviewer
Pass
Audited by Gen Agent Trust Hub on Jun 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions define clear, benign behavior for reviewing work plans provided by users. While the skill processes external input and has the capability to read local files for verification, these actions are essential to its primary purpose and do not involve unauthorized data exfiltration, command execution, or other malicious patterns.
- [PROMPT_INJECTION]: The skill processes user-provided work plans, which constitutes an ingestion point for untrusted data (Indirect Prompt Injection surface). Given the skill's role as a review tool, this is considered a standard functional surface.
- Ingestion points: User-provided plans passed to the reviewer (SKILL.md).
- Boundary markers: None defined to isolate user content from instructions.
- Capability inventory: Filesystem access to 'open referenced files' for verification purposes (SKILL.md).
- Sanitization: No explicit validation of plan content is specified.
Audit Metadata