Vmware Aiops

Use this skill whenever the user needs to manage VMs in VMware/vSphere/ESXi — it's the entry point for all VM operations. Directly handles: power on/off, clo...

This skill appears internally consistent with VMware VM management. Before installing: 1) Verify the package source and checksum on PyPI/GitHub (registry shows 'Source: unknown'); 2) Review the open-source repo (vmware-aiops) to confirm prompt-injection protections, logging behavior, and that passwords are never leaked; 3) Use least-privilege vCenter/ESXi service accounts for the per-target passwords stored in ~/.vmware-aiops/.env and set strict file permissions (chmod 600); 4) Be cautious enabling guest-exec or daemonized scanning in production (these can access VM guest output and host logs); 5) Only enable webhooks to endpoints you control and audit the webhook payload format; 6) Consider testing in an isolated lab before production and pin the tool version when installing from PyPI.

Type: OpenClaw Skill Name: vmware-aiops Version: 1.5.15 The skill provides high-privilege administrative control over VMware infrastructure, including the ability to execute arbitrary commands inside guest VMs (vm_guest_exec), upload/download files, and perform destructive batch operations. While these capabilities are aligned with the stated purpose and the documentation highlights safety features like audit logging (via vmware-policy) and double-confirmation for destructive actions, the inherent risk of Remote Code Execution (RCE) and credential management within an AI agent context is significant. Additionally, the inclusion of a background daemon for log scanning and webhook integration (Slack/Discord) in SKILL.md and references/capabilities.md increases the potential attack surface for data exfiltration or unauthorized monitoring if the agent is compromised.