start
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill reads project-specific metadata and documentation (e.g.,
.agents/docs/technical-preferences.md) to suggest an appropriate workflow. This context gathering is internal and does not involve network exfiltration or access to sensitive global credentials. - [SAFE]: The skill writes local configuration files to the
production/directory to manage project stage and review preferences. These writes are confined to the current project and do not impact system-level files or persistence mechanisms. - [SAFE]: Interactive branching is handled via
AskUserQuestion, ensuring that the user selects their preferred path. The skill's hand-off process is transparent, instructing the user on which subsequent command to execute manually.
Audit Metadata