de-ai-pipeline

Pass

Audited by Gen Agent Trust Hub on Mar 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a deterministic workflow for document cleanup. It relies on a local script (scripts/scan-ai-patterns.py) and a local JSON database for pattern matching, which is a standard development practice for automated linting or formatting.
  • [SAFE]: File system interactions are limited to reading documentation, using standard grep/glob tools, and performing targeted edits. The skill follows best practices by staging changes with git add but explicitly avoiding automatic commits.
  • [SAFE]: The skill does not perform any network operations, does not handle sensitive credentials, and contains no obfuscated code or hidden instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 23, 2026, 04:36 PM
Security Audit — agent-trust-hub — de-ai-pipeline