dark-mode-testing
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill navigates to external websites and processes their content using
browser_snapshotand screenshots. This creates a surface where instructions hidden on a website (e.g., in text or metadata) could potentially influence the agent's actions. - Ingestion points: Website content retrieved via
browser_snapshotand visual data frombrowser_take_screenshotinSKILL.md. - Boundary markers: No explicit instructions are provided to the agent to ignore or delimit instructions found within the processed web content.
- Capability inventory: The skill uses
browser_navigate,browser_take_screenshot,browser_snapshot, and executes JavaScript within the browser context to toggle theme attributes. - Sanitization: There is no evidence of sanitization or filtering of the external content before it is processed by the agent.
Audit Metadata