Loading ToolBench
Chat with your Proxmox VE cluster. A Model Context Protocol (MCP) server that lets any LLM list nodes, VMs and LXC containers, read status, manage lifecycle and take snapshots — safely.
35 tools with mixed quality. Naming is strong (verb_noun pattern, clear intent). Descriptions present but often generic (avg ~80 chars, below 194-char baseline). Critical gap: input schemas are inferred from smoke.mjs test file, not visible in actual tool registration code. Cannot verify parameter types, constraints, or output schemas. Many tools lack parameter descriptions. No evidence of error handling guidance, recovery paths, or confirmation patterns for destructive operations. Composition is reasonable (single responsibility per tool), but chaining support unclear.
Information about the Proxmox MCP Server, including version, author, and support links
Create a backup of a VM or container
Clone a VM or container
Check the health status of the Proxmox cluster, including quorum, node status and replication
Get an overview of cluster-wide resource usage: total CPU, memory, disk and guest counts
Create a new LXC container
Input schemas not visible in source code. Tool definitions inferred from smoke.mjs test file only. Cannot verify parameter types, constraints, enums, or descriptions.
No error handling guidance visible. Destructive operations (delete_guest, rollback_snapshot, restore_backup, stop_guest) lack confirmation patterns or recovery instructions. LLM cannot determine if errors are retryable or what to do next.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 52 | <=2025-11-25 | v2 |
Create a snapshot of a VM or container
Create a new QEMU VM from a template or from scratch
Delete a VM or container permanently
Delete a snapshot of a VM or container
Find VMs and containers with low CPU and memory usage
Find unused disk volumes in storage
Get OS information and guest agent details for a VM or container
Get detailed status and configuration of a specific VM or container
List all available backups for a guest or storage
List all QEMU VMs and LXC containers in the cluster with their status, resource usage, and configuration
List all nodes in the Proxmox cluster with their status, uptime, CPU, memory and disk usage
List all generated resilience and compliance reports
List all snapshots of a VM or container
List all storage devices and their usage in the cluster
List recent tasks and their status in the cluster
List available VM and container templates
Migrate a VM or container to another node
Orchestrate patching of nodes with automatic rollback capability
Reboot a running VM or container
Restore a VM or container from a backup
Resume a suspended VM
Rollback a VM or container to a previous snapshot
Run a disaster recovery drill and generate evidence report
Modify CPU, memory or disk resources of a guest
Gracefully shut down a running VM or container
Start a stopped VM or container
Force stop a running VM or container (pulls the virtual power cord)
Suspend a running VM (saves state to disk)
Verify the integrity of backups and generate a compliance report
Parameter descriptions missing or minimal. E.g., 'vmid' described as 'The numeric VMID of the guest' (generic); no guidance on valid ranges, format, or how to obtain VMID. Descriptions average ~80 chars vs 194-char baseline.
No output schema documentation visible. LLM cannot plan downstream tool calls or extract required fields (e.g., does list_guests return vmid, name, status, resource usage?). Cannot verify if chaining IDs are included.
STDIO transport only. Not remotely accessible; cannot be used by hosted MCP clients or cloud-based agents. Requires local process spawning.