conf-preference-elicitation
Pass
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a set of instructional prompts and logic for personalizing conference schedules. No command execution, network exfiltration, or credential access patterns were found.
- [SAFE]: The workflow involves reading local JSON data (clusters, affinities) and writing a profile JSON based on user interaction. There are no signs of remote code execution or unauthorized file access.
- [SAFE]: The instructions emphasize user interaction and human checkpoints, explicitly advising against background execution and over-automation.
Audit Metadata