planning-under-uncertainty

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFE
Full Analysis
  • Prompt Injection (SAFE): No instructions designed to bypass agent safety filters or override system prompts were detected. The skill uses standard instructional language for structured planning.
  • Data Exposure & Exfiltration (SAFE): No hardcoded credentials, sensitive file paths, or unauthorized network operations were found. The skill only produces text-based planning documents.
  • Unverifiable Dependencies & RCE (SAFE): There are no package installation commands (npm/pip) or patterns of remote code execution (curl/wget | bash).
  • Indirect Prompt Injection (LOW): The skill ingests untrusted user context about projects and initiatives (SKILL.md). While this represents a surface for injection, the skill's capabilities are limited to text generation and file writing of Markdown templates, with no path to command execution or sensitive data access.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 03:52 PM