AGENT LAB: SKILLS

x-article-publisher

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE] (SAFE): The scripts read local files (images, HTML, markdown) provided via command-line arguments. This is the intended functionality for these tools. No unauthorized data access or exfiltration patterns were found.
  • [COMMAND_EXECUTION] (SAFE): The scripts do not use subprocesses to execute shell commands. They interact with system APIs through established libraries such as AppKit on macOS and win32clipboard on Windows.
  • [EXTERNAL_DOWNLOADS] (SAFE): There are no network requests or remote code downloads in the code. Dependencies are listed in documentation for manual installation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:09 PM