AI agents invoke pve_shutdown_vm to trigger actions in Proxmox. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
Shutting down a VM/container is an operational action that stops running workloads. It is not destructive (data is preserved, VM can be restarted), but it executes an external operation with real-world impact — potentially causing service outages. This places it firmly in Execute. The blast radius is high because misuse could bring down production systems.
From the tool's definition Gracefully shut down a VM/container. Requires confirm=true.
Attacks that exploit this kind of access
Gracefully shut down a VM/container. Requires confirm=true. It is categorised as a Execute tool in the Proxmox MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Proxmox MCP server in PolicyLayer and add a rule for pve_shutdown_vm: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Proxmox. Nothing to install.
pve_shutdown_vm is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the pve_shutdown_vm rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for pve_shutdown_vm. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
pve_shutdown_vm is provided by the Proxmox MCP server (plgonzalezrx8/proxmox-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →