skills/niach/manyclaws/manyclaws-k8s/Gen Agent Trust Hub

manyclaws-k8s

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: Employs kubectl to perform cluster operations such as applying resource manifests, scaling workloads, and inspecting pod logs within the user's namespace.
  • [COMMAND_EXECUTION]: Uses curl to interact with an internal service (manyclaws-controller.manyclaws-system.svc) for managed cross-namespace deployment operations.
  • [SAFE]: Authentication for the internal API is securely handled via an environment variable ($MANYCLAWS_API_TOKEN) rather than hardcoded credentials.
  • [SAFE]: The skill instructions include explicit platform restrictions, such as the inability to access other namespaces or manage secrets directly, minimizing the risk of unauthorized privilege escalation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 08:13 PM
Security Audit — agent-trust-hub — manyclaws-k8s