workflow-onboarding

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the GitHub CLI (gh) to manage repository labels. It performs read operations to detect current state and write operations to create or update labels. These actions are restricted to a predefined list of six canonical labels related to workflow onboarding.
  • [EXTERNAL_DOWNLOADS]: The skill fetches label definitions and documentation templates from the vendor's own repository (dividedby/skills). These are recognized as trusted vendor resources used for configuration synchronization.
  • [SAFE]: Implements a robust Human-In-The-Loop (HITL) gate. The instructions explicitly require the agent to present a full execution plan and wait for unambiguous user confirmation (e.g., 'go', 'approved') before proceeding with any network mutations or file writes.
  • [SAFE]: Included Python script (loop-labels.test.py) is a pure utility for parsing markdown documentation and verifying the integrity of the label install plan. It relies solely on Python standard libraries (re, unittest, pathlib) and does not perform network operations or unsafe dynamic execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 05:41 AM
Security Audit — agent-trust-hub — workflow-onboarding