task-metadata-audit

Pass

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is strictly limited to read-only operations on local Markdown files within the workspace knowledge directory. It does not include any tools or instructions for network communication, external data exfiltration, or command execution.
  • [SAFE]: No obfuscation techniques, hardcoded credentials, or sensitive file path exposures were detected.
  • [SAFE]: The skill includes explicit guardrails that prevent unauthorized file modifications or Kanban board movements, requiring human-in-the-loop approval for any suggested remediation.
  • [SAFE]: All external references are local to the repository and follow standard documentation practices.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 14, 2026, 05:11 AM
Security Audit — agent-trust-hub — task-metadata-audit