plan-feature
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Indirect Prompt Injection] (SAFE): The skill is designed to ingest and analyze external data such as user requirements and existing codebase files. This creates a surface for indirect prompt injection. However, per the instructions to consider the skill's primary purpose, this risk is minimal as the skill explicitly forbids writing code ('DO NOT WRITE CODE') and acts solely as a documentation and planning aid.
- [Data Exposure] (SAFE): The Analysis phase involves locating relevant files and mapping codebase patterns. This involves reading local files, but the skill lacks network capabilities or exfiltration patterns, and the activity is strictly aligned with its stated purpose of feature planning.
Audit Metadata