browser-workflow-generator
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains instructional markers such as 'CRITICAL' and 'IMPORTANT' which are used exclusively for operational logic and task management rather than attempting to override AI safety constraints.
- [SAFE]: The skill uses specialized sub-agents for exploration and research, which is a standard orchestration pattern for complex tasks and is contained within the agent's internal tool environment.
- [SAFE]: A robust security control is implemented in Phase 7, requiring explicit user approval and a review of the generated content before the file is written to the disk.
- [SAFE]: Codebase exploration and web search activities are limited to their stated purposes (UX research and feature discovery) and do not involve the execution of untrusted external scripts or the exfiltration of sensitive data.
Audit Metadata