nw-roadmap-design

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a set of purely instructional principles and quality gates for creating efficient software implementation roadmaps, emphasizing token efficiency and separation of concerns between design and implementation.
  • [PROMPT_INJECTION]: The skill includes logical constraints such as the 'Measure Before Plan Gate' and 'Simplest Solution Gate.' These require the agent to verify data availability or document alternatives before proceeding, serving as workflow controls rather than attempts to bypass agent safety protocols.
  • [DATA_EXFILTRATION]: While the skill instructs the agent to read local project files (e.g., tests/acceptance/test_*.py) to map roadmap steps to scenarios, it does not include any mechanisms for exfiltrating this data or accessing sensitive system directories (like .ssh or .aws).
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 06:45 PM
Security Audit — agent-trust-hub — nw-roadmap-design