Proxmox Aiops

Use this skill whenever the user needs to manage VMs and containers on Proxmox VE — list/inspect/configure VMs, power and lifecycle (start/stop/shutdown/reboot/reconfigure/clone/delete/migrate), snapshots (create/delete/list/rollback), LXC containers (list/start/stop), cluster/node status, async task polling, and storage listing. Always use this skill for "list proxmox vms", "start proxmox vm", "stop proxmox vm", "proxmox snapshot", "proxmox vm status", "migrate proxmox vm", "proxmox container", or "list proxmox storage" when the context is explicitly Proxmox / Proxmox VE / PVE. Do NOT use for non-Proxmox hypervisors, Kubernetes, or cloud providers. Preview — broad coverage of common Proxmox operations, with a built-in governance harness (audit, policy, token budget, undo, risk-tiers).

Proxmox AIops (preview)

Disclaimer: This is a community-maintained open-source project and is not affiliated with, endorsed by, or sponsored by Proxmox Server Solutions GmbH. "Proxmox" is a trademark of its owner. Source code is publicly auditable at github.com/AIops-tools/Proxmox-AIops under the MIT license.

Governed VM and container lifecycle operations for Proxmox VE — 23 MCP tools, every one wrapped with the bundled @governed_tool harness: a local unified audit log under ~/.proxmox-aiops/, policy engine, token/runaway budget guard, undo-token recording, and graduated-autonomy risk tiers.

Standalone: the governance harness is bundled in the package (proxmox_aiops.governance) — proxmox-aiops has no external skill-family dependency. Preview: broad coverage of common Proxmox operations, not yet exhaustive.

What This Skill Does

Category	Tools	Count	Read or Write
VM Lifecycle	list, get, config, start, stop, shutdown, reboot, reconfigure, clone, delete, migrate	11	3 read / 8 write
Snapshots	create, delete, list, rollback	4	1 read / 3 write
LXC Containers	list, start, stop	3	1 read / 2 write
Cluster / Tasks	node list, cluster status, async task poll	3	3 read
Storage	list pools, list content	2	2 read

Quick Install

uv tool install proxmox-aiops
proxmox-aiops doctor

When to Use This Skill

• List/inspect Proxmox QEMU VMs and their config
• Power ops: start, hard-stop, graceful shutdown, reboot
• Reconfigure (cores/memory), clone, delete, or migrate a VM between nodes
• Create / delete / list / roll back VM snapshots
• Manage LXC containers (list, start, stop)
• Inspect cluster nodes, quorum, and poll async tasks (clone/migrate) by UPID
• List storage pools and their content (ISOs, disk images, backups)

Do NOT use when the target is not Proxmox VE (other hypervisors, Kubernetes, or cloud providers are out of scope for this skill).

Common Workflows

Snapshot, then reconfigure a VM

1. proxmox-aiops vm list → find the vmid and confirm it is the right VM/node
2. proxmox-aiops vm snapshot-create <vmid> --name pre-change → baseline before any risky change
3. Make the change (outside this skeleton's scope for now)
4. Failure branch: if the change goes wrong, the vm_snapshot_create MCP call recorded an _undo_id (look it up with proxmox-aiops audit tooling), then run the inverse — e.g. proxmox-aiops vm snapshot-rollback <vmid> --name pre-change, or vm snapshot-delete to clean up after a manual recovery.

Stop a VM safely

1. proxmox-aiops vm get <vmid> → confirm current status is running
2. proxmox-aiops vm stop <vmid> --dry-run → preview the exact API call
3. proxmox-aiops vm stop <vmid> → double confirmation required; vm_stop records an inverse vm_start undo descriptor
4. Failure branch: if doctor shows the node unreachable or the secret env var is missing, fix ~/.proxmox-aiops/.env (chmod 600) before retrying — the stop is never issued against an unauthenticated session.

Usage Mode

Scenario	Recommended	Why
Local/small models (Ollama, Qwen)	CLI	fewer tokens than MCP
Cloud models (Claude, GPT)	Either	MCP gives structured JSON I/O
Automated pipelines	MCP	type-safe parameters, audited

MCP Tools (23 — 8 read, 15 write)

Category	Tools	R/W
VM Lifecycle	vm_list, vm_get, vm_config	Read
	vm_start, vm_stop, vm_shutdown, vm_reboot, vm_reconfigure, vm_clone, vm_delete, vm_migrate	Write
Snapshots	vm_list_snapshots	Read
	vm_snapshot_create, vm_snapshot_delete, vm_snapshot_rollback	Write
LXC Containers	ct_list	Read
	ct_start, ct_stop	Write
Cluster / Tasks	node_list, cluster_status, task_status	Read
Storage	storage_list, storage_content	Read

Harness features that light up: write tools with a clean inverse (vm_start/vm_stop/vm_shutdown/vm_reconfigure/vm_clone/vm_migrate/vm_snapshot_create/ct_start/ct_stop) pass an undo= lambda so the harness records an inverse descriptor (with _undo_id) to the undo store — vm_reconfigure captures the prior cores/memory, vm_clone's inverse is vm_delete(newid), vm_migrate's is migrate-back. Irreversible writes (vm_delete, vm_snapshot_rollback) declare no undo and are tagged risk_level=high. All 23 tools are audit-logged under ~/.proxmox-aiops/ and pass through the policy pre-check + budget/runaway guard + graduated risk-tier gate. Proxmox writes are async (return a task UPID) — poll with task_status instead of re-issuing (the runaway breaker backs this up).

CLI Quick Reference

proxmox-aiops vm list [--target <t>] [--node <n>]
proxmox-aiops vm get <vmid> [--node <n>]
proxmox-aiops vm start <vmid> [--node <n>]
proxmox-aiops vm stop <vmid> [--dry-run]              # double confirm
proxmox-aiops vm snapshot-create <vmid> --name <snap>
proxmox-aiops vm snapshot-delete <vmid> --name <snap> [--dry-run]   # double confirm
proxmox-aiops vm snapshot-list <vmid>
proxmox-aiops storage list [--node <n>]
proxmox-aiops doctor
proxmox-aiops mcp                                      # start MCP server (stdio)

Troubleshooting

"Config file not found"

Create ~/.proxmox-aiops/config.yaml with a targets: list (see README), and put secrets in ~/.proxmox-aiops/.env (chmod 600).

"Secret not found. Set environment variable: PROXMOX__SECRET"

Each target needs a per-target secret env var. For target pve-lab, set PROXMOX_PVE_LAB_SECRET=<token-uuid> in .env.

"Token auth requires user in the form 'user@realm!tokenid'"

For API-token auth (recommended, least privilege), user must include the token id after !, e.g. root@pam!claude. For password auth set auth_kind: password and use user@realm.

"No node specified and no default node configured"

Either pass --node <name> / node=<name>, or set node: on the target in config.yaml. VM operations can auto-locate a vmid across nodes, but storage listing needs an explicit node.

Audit & Safety

All operations are automatically audited via the bundled @governed_tool decorator (proxmox_aiops.governance):

• Every tool call logged to ~/.proxmox-aiops/audit.db (local SQLite audit DB; relocate with PROXMOX_AIOPS_HOME)
• Policy rules enforced via ~/.proxmox-aiops/rules.yaml (deny rules, maintenance windows, risk tiers)
• Budget / runaway guard caps cumulative tool calls and wall-time, and trips on tight poll/retry loops
• Undo store records inverse descriptors for reversible writes (start/stop/shutdown/reconfigure/clone/migrate/snapshot-create, container start/stop)
• Graduated-autonomy risk tiers gate write operations (require a recorded approver for the highest tiers)

The harness is bundled in the package — no external dependency, no manual setup.

License

MIT — github.com/AIops-tools/Proxmox-AIops