enterprise-code-architect

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOW
Full Analysis
  • [Indirect Prompt Injection] (INFO): The skill utilizes the 'Read' tool to access local reference files within its own directory. While this is a data ingestion surface, it is used only for static internal documentation and does not process untrusted external content.
  • [Data Exposure] (SAFE): No access to sensitive system paths, environmental variables, or hardcoded credentials was found.
  • [Command Execution] (SAFE): The skill instructions are limited to read operations and do not attempt to execute shell commands, scripts, or subprocesses.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 01:10 PM