mage-ad-localizer
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill processes user-provided images and text but mitigates indirect prompt injection risks through a 'Human-in-the-loop' design. It requires the agent to present all translated copy to the user for approval before invoking the generation tool.
- [SAFE]: All external URLs and tools (such as the Mage MCP connector) belong to the skill's verified author, mage-space. The instructions for setting up the connector follow standard procedures for the Model Context Protocol (MCP).
- [SAFE]: No obfuscation, persistence mechanisms, or unauthorized privilege escalation attempts were detected. The skill's operations are confined to its stated purpose of ad localization.
Audit Metadata