browser-edge-cases
Pass
Audited by Gen Agent Trust Hub on Jul 9, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill inherently processes untrusted data by ingesting HTML content, snapshots, and console logs from external websites during automation tasks. While this represents a potential surface for indirect prompt injection, it is the primary and intended function of a browser automation skill. The provided scripts focus on technical debugging and structural analysis (e.g., detecting Shadow DOM or scrollable containers) rather than automated decision-making based on semantic content.
- [COMMAND_EXECUTION]: The diagnostic scripts use a browser bridge to execute JavaScript within the browser context. This is standard behavior for advanced browser automation and is used here for legitimate purposes such as calculating element visibility, detecting frameworks (React/Vue), and traversing the DOM. No arbitrary shell command execution or system-level exploitation was identified.
- [EXTERNAL_DOWNLOADS]: The skill references well-known services and websites (e.g., LinkedIn, X/Twitter, example.com, and W3Schools) for testing purposes. These are established domains used to verify automation capabilities against real-world complexity and do not involve downloading or executing untrusted code from malicious sources.
Audit Metadata