visual-ralph

Pass

Audited by Gen Agent Trust Hub on Apr 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to run the application, capture screenshots, and execute repository-native build, lint, and test scripts to verify the implementation matches the visual target.\n- [EXTERNAL_DOWNLOADS]: The skill allows for the ingestion of visual data from user-provided live URLs to establish a baseline for implementation, which acts as a data ingestion point for the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 26, 2026, 06:47 AM