Proxmox Agent Lab Orchestrator v1.0

Price: $4.99

Category: orchestration

Status: Ready for listing

Viper QA: APPROVED_FOR_RELEASE on 2026-05-08; 7/7 tests passed; Security Audit: CLEARED by Warlock.

What it does

Proxmox Agent Lab Orchestrator turns a planned OpenClaw, Hermes, browser-agent, or coding-agent workload into a safe Proxmox lab design before anyone touches a live cluster. It recommends VM/LXC groups, per-agent resource allocation, backup windows, isolation tiers, and operational risk flags so buyers can size agent infrastructure with confidence.

It is deliberately read-only: the skill generates a planning report and implementation checklist, but does not connect to Proxmox, create virtual machines, alter networking, or delete snapshots.

Use cases

AI operators building persistent agent labs who need clear VM/LXC sizing, storage reserves, and backup guidance before standing up long-running agents.
Consultants and SMB teams deploying OpenClaw infrastructure who want a repeatable planning template for homelab-to-production Proxmox environments.
Teams running browser or coding agents who need stronger isolation recommendations for heavier workloads without overbuilding the whole cluster.

Requirements

OpenClaw v2026.3.23 or newer recommended.
Python 3.10+.
No external Python packages required; uses only the Python standard library (argparse, json, math, sys).
No Proxmox API key required for planning mode.
Before implementing the generated plan on a real cluster, operators must independently verify Proxmox node names, bridges, storage pools, credentials, snapshots, and restore paths.

Example usage


python3 skills/proxmox-agent-lab-orchestrator/scripts/proxmox_agent_lab_plan.py \
  --agents 8 \
  --profile mixed \
  --ram-gb 64 \
  --cpu-cores 16 \
  --storage-gb 1000 \
  --backup-window "02:00-04:00 local" \
  --format markdown

Expected output


# Proxmox Agent Lab Plan — PASS

## Groups
- 6x LXC `agent-core-lxc`: 4GB RAM, 2 vCPU, 50GB disk each (standard isolation)
- 2x VM `browser-or-build-vm`: 4GB RAM, 2 vCPU, 60GB disk each (strong isolation)

## Findings
- INFO profile: coding agents, browser-lite, scheduled jobs

## Checklist
- Confirm Proxmox node, bridge, DNS, and storage pool names.
- Create golden template with OpenClaw prerequisites but no secrets.
- Use per-agent credentials and firewall rules.
- Schedule backups during 02:00-04:00 local and test one restore.
- Snapshot before upgrades; keep an out-of-band recovery shell.

When capacity is too low, the command returns a FAIL status, exits non-zero, and explains the missing RAM or storage reserve. CPU overcommit is reported as WARN so operators can make an informed decision.

Security and QA notes

Viper approved the skill for release on 2026-05-08 after 7/7 tests passed. The security gate passed with Warlock clearance, and the implementation remains bounded/read-only: no network calls, no subprocess execution, no Proxmox mutation, and no live cluster actions.

Website card copy

Design a safer Proxmox lab for long-running AI agents before you build it. Proxmox Agent Lab Orchestrator converts your agent count, workload profile, RAM, CPU, storage, and backup window into a practical VM/LXC plan with isolation tiers, capacity reserves, risk flags, and an implementation checklist — all read-only, with no live cluster changes.

Proxmox Agent Lab Orchestrator