Vmware Storage

Use this skill whenever the user needs to manage VMware storage — datastores, iSCSI targets, and vSAN clusters. Directly handles: browse datastores, scan for...

This package mostly looks like a legitimate VMware storage CLI, but there are several inconsistencies you should resolve before installing or running it on production systems: 1) Confirm the correct install method (SKILL.md lists a 'uv' installer and pip/source options, but registry metadata claims no install spec). Prefer installing from a trusted source (official GitHub repo or your corporate package mirror) and review the actual package before installation. 2) Verify where audit logs are written (SKILL.md mentions ~/.vmware/audit.db but other paths use ~/.vmware-storage) and ensure the audit DB location is acceptable and backed up if needed. 3) Clarify the "local-only / no webhooks" claim — the tool documents an optional webhook_url and the doctor command makes outbound network checks and vSphere API calls; if you do not want outbound webhooks, ensure webhook_url is empty and network egress to untrusted endpoints is blocked. 4) Inspect the vmware-policy dependency and any shared config (~/.vmware/rules.yaml) to understand policy checks and ensure they won't unintentionally affect other tools. 5) Before granting it access, ensure your ~/.vmware-storage/.env file permissions are 600 and never store plaintext passwords elsewhere (SKILL.md recommends this). 6) If you need higher assurance, fetch the source from the referenced GitHub (https://github.com/zw008/VMware-Storage) and audit the package code and the vmware-policy dependency for unexpected behavior (web requests, uploads, or unexpected third-party hosts).

Type: OpenClaw Skill Name: vmware-storage Version: 1.5.15 The vmware-storage skill is a legitimate tool for managing VMware datastores, iSCSI, and vSAN clusters via the vSphere API. It incorporates several security best practices, including local-only stdio transport, mandatory audit logging to a local SQLite database via the vmware-policy dependency, and a dedicated _sanitize() function to prevent prompt injection from malicious file names on datastores. No evidence of data exfiltration, obfuscation, or unauthorized execution was found across the documentation or tool definitions (SKILL.md, setup-guide.md).