architecture

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Overall Assessment] (SAFE): The skill consists entirely of Markdown-based guidelines, templates, and reference materials. It does not include any scripts (Python, JavaScript, Shell) or configurations that could lead to code execution.
  • [Tool Usage] (SAFE): The allowed tools are limited to Read, Glob, and Grep, which are used to navigate and read the provided documentation files. There are no tools for network access or file modification.
  • [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths, or data exfiltration patterns were found. The skill does not perform any network operations.
  • [Indirect Prompt Injection] (LOW): While the skill interacts with user-provided requirements to generate architecture suggestions, it does so within a structured framework and does not ingest data from untrusted external sources like web pages or third-party APIs.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:48 PM