plan-review
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): The instructions use emphasized language like 'CRITICAL' and 'IMPORTANT' to direct agent behavior within its operational environment, rather than attempting to bypass safety filters or jailbreak the underlying model.
- [Data Exposure & Exfiltration] (SAFE): The skill is designed to read local planning files (plan.md, phase-*.md). No network tools or exfiltration patterns are utilized.
- [Indirect Prompt Injection] (LOW): The skill ingests untrusted data from project files. While an attacker could attempt to hide instructions in a 'plan.md', this skill does not invoke dangerous tools (like code execution or shell access) that would allow such an injection to escalate permissions or exfiltrate data.
- [Remote Code Execution] (SAFE): No remote scripts, package installations, or dynamic execution patterns were found.
Audit Metadata