agent-primitives
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a documentation-only resource that provides design principles and skeletons for organizing agent primitives. It does not contain any executable scripts or active commands.
- [SAFE]: The instructions explicitly promote security best practices, including the 'four safety pillars': least privilege, sandboxing, rate limits, and audit logs.
- [SAFE]: No patterns of prompt injection, data exfiltration, obfuscation, or unauthorized privilege escalation were detected in either the skill body or the reference files.
Audit Metadata