proxmox-aiops
Related Servers
Alternatives to proxmox-aiops
No user-submitted related servers found.
Related Servers
- AlicenseNot gradedqualityDmaintenanceMCP server that lets an AI agent manage Proxmox VE in natural language, with policy-driven security, read-only mode, two-step confirmations, and audit logging.MIT
- FlicenseNot gradedqualityDmaintenanceMCP server for Proxmox VE that enables AI assistants to inspect and manage LXC containers, VMs, snapshots, and resource pools via the Proxmox API.-
- AlicenseBqualityCmaintenanceA Model Context Protocol (MCP) server for Proxmox Virtual Environment that enables AI assistants to manage virtual machines, containers, nodes, and resources through natural language interactions.104MIT
- AlicenseAqualityDmaintenanceMCP server for managing Proxmox VE resources, VM/CT lifecycle, snapshots, backups, and clones via natural language.23MIT
- AlicenseNot gradedqualityDmaintenanceMCP server for full Proxmox VE management - VMs, containers, storage, backups, networking.3MIT
- AlicenseNot gradedqualityCmaintenanceEnterprise MCP server for Proxmox VE, offering 298 tools to manage VMs, containers, storage, cluster, firewall, and more via natural language.1MIT
TDQS
Scored across 43 tools
Most tools target a clearly distinct resource+action (vm_list vs vm_get vs vm_config, snapshots vs backups vs storage, node/cluster/HA/task families). The only real overlap is among the power tools (vm_stop vs vm_shutdown vs vm_reboot) and the various RCA/inventory reads (cluster_resources vs vm_list vs node_status), but the descriptions explicitly contrast these, so an agent can choose correctly.
Strong resource-prefix scheme throughout (vm_*, ct_*, node_*, cluster_*, task_*, backup_*, pool_*, storage_*, ha_*, undo_*) with all snake_case. There are minor verb-order deviations (vm_list vs node_status, vm_backup vs backup_list) but the pattern stays predictable and readable.
43 tools is well above the usual sweet spot and forces the agent to scan a large surface, though the breadth of Proxmox (VMs, CTs, nodes, cluster, storage, backups, HA, firewall, undo) means each tool maps to a real operation rather than bloat. It sits at the heavy end of appropriate.
Lifecycle coverage is strong: inventory, start/stop/shutdown/reboot, migrate, resize/move disk, snapshots with rollback, backup/restore, tasks, undo, plus read-only cluster/storage/HA/firewall views. Gaps exist — no VM/CT creation from scratch (clone only), backup_restore is QEMU-only, and firewall/pool tools are read-only — but agents can work around most of these.