stitch-code-to-design

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill workflow is restricted to specific MCP tools and skill-specific local directories, maintaining clear operational boundaries.
  • [SAFE]: Includes explicit instructions in the 'Anti-Patterns' section to prevent the agent from reading, storing, or committing sensitive credentials like API keys or cookies.
  • [SAFE]: Implements a human-in-the-loop requirement for uploading artifacts, ensuring the user approves all data sent to the remote service.
  • [SAFE]: All external references for support materials and updates point to official and reputable repositories affiliated with the platform provider.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 06:06 PM
Security Audit — agent-trust-hub — stitch-code-to-design