proxmox-system-administration
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for using official Proxmox CLI tools (qm, pct, pvesh, pvesm) to manage virtual machines and containers. All provided examples use sanitized placeholders, ensuring no sensitive host information is leaked or hardcoded.
- [INDIRECT_PROMPT_INJECTION]: The skill involves reading system state from the Proxmox hypervisor. This creates a surface where external environmental data (such as VM names or storage statuses) is brought into the agent context. However, the skill implements strong logical boundaries (e.g., root/guest/account boundaries) and verification steps to manage this surface.
- Ingestion points: Command output from hypervisor discovery tools in SKILL.md.
- Boundary markers: Includes an Operating Policy and Verification Checklist to guide safe execution.
- Capability inventory: Uses shell commands for hypervisor management and guest user creation.
- Sanitization: All variable inputs are treated as placeholders (e.g., , ), requiring manual user replacement.
Audit Metadata