inspector-overlay
Pass
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a development-only UI component intended to assist developers in providing context to AI agents. It does not perform any unauthorized background actions.
- [SAFE]: Data collection is limited to DOM element metadata (tags, selectors, text previews, and bounding boxes) and is only triggered by explicit user interaction (holding the Alt key and clicking).
- [DATA_EXFILTRATION]: The skill facilitates the collection of codebase metadata, such as file paths and component names via 'data-file' and 'data-component' attributes. This information is placed on the user's system clipboard for manual pasting into an AI agent and is not transmitted over the network to any third party. This aligns with the skill's primary documented purpose.
- [SAFE]: The implementation in 'LlmInspectorOverlay.tsx' includes a default guard ('process.env.NODE_ENV === "development"') to prevent the overlay from being active in production environments, following security best practices for development tools.
- [SAFE]: No obfuscation, remote code execution patterns, or suspicious network operations were found in the provided files.
Audit Metadata