Skip to content
@AIops-tools

AIops-tools

AIops-tools

Governed AI operations tooling for infrastructure — audit, budget, undo, and graduated approval built into every tool.

AI agents are great at infrastructure ops, right up until something goes wrong in production. AIops-tools is a family of MCP servers + CLIs that wrap real infrastructure platforms with a governance harness, so an agent's actions are:

  • Audited — every operation logged to a local SQLite trail (who / what / why / when), secret-redacted.
  • Bounded — per-process token/call budget + a runaway-loop breaker (no more "one operation, 26k tokens").
  • Reversible — write operations record an inverse undo token wherever a clean inverse exists.
  • Graduated — risk tiers gate writes by environment/tag; the highest tiers require a recorded approver.
  • Safe by default — destructive ops need double confirmation + --dry-run; all API text is sanitized.

Every tool is self-contained (the harness is bundled — no shared runtime dependency) and ships on PyPI, the MCP Registry, and ClawHub.

Published tools

Tool Platform Install MCP tools
proxmox-aiops Proxmox VE — VMs, LXC, snapshots, cluster, storage, backups, HA, pools, firewall pip install proxmox-aiops 39
veeam-aiops Veeam Backup & Replication — jobs, restore, repositories, sessions, infrastructure pip install veeam-aiops 21
k8s-aiops Kubernetes — k3s / EKS / GKE / AKS (workloads, batch, config, storage, networking, rollouts) pip install k8s-aiops 51
network-aiops Network devices via NAPALM — Cisco IOS/NX-OS/IOS-XR, Arista EOS, Juniper Junos + NetBox pip install network-aiops 28
truenas-aiops TrueNAS SCALE storage — pools, datasets, snapshots, disks, alerts, services pip install truenas-aiops 21
ot-aiops Industrial / OT data tap — OPC-UA, Modbus, S7comm, Mitsubishi MC, MTConnect, MQTT/Sparkplug B, EtherNet/IP, EtherCAT + diagnostics, OEE/downtime & asset inventory pip install ot-aiops 57

How it works

Each tool exposes both a CLI and an MCP server (<tool> mcp, stdio transport). Point any MCP client (Claude, etc.) at it. Configuration and the audit/undo/policy store live under ~/.<tool>/ (relocatable via <TOOL>_AIOPS_HOME). Credentials are kept in an encrypted store (secrets.enc, Fernet + master password) — never in plaintext. A friendly <tool> init wizard walks you through first-time setup.

pip install proxmox-aiops          # or pipx / uv tool install
proxmox-aiops init                 # interactive setup (encrypts your credentials)
proxmox-aiops doctor               # verify connectivity
proxmox-aiops mcp                  # run as an MCP server

License & affiliation

MIT licensed. Community-maintained. Not affiliated with, endorsed by, or sponsored by the vendors of the platforms these tools operate (Proxmox Server Solutions GmbH, Veeam Software, the CNCF / Kubernetes project, etc.); all trademarks belong to their respective owners.

Popular repositories Loading

  1. Proxmox-AIops Proxmox-AIops Public

    Standalone governed Proxmox VE ops — 23 MCP tools with built-in audit/budget/undo/risk-tier harness

    Python 1

  2. Veeam-AIops Veeam-AIops Public

    Standalone governed Veeam Backup & Replication ops — 12 MCP tools with built-in audit/budget/undo/risk-tier harness

    Python

  3. .github .github Public

    AIops-tools org profile

  4. K8s-AIops K8s-AIops Public

    Standalone governed Kubernetes ops — 15 MCP tools with built-in audit/budget/undo/risk-tier harness

    Python

  5. Network-AIops Network-AIops Public

    Standalone governed multi-vendor network ops (NAPALM) — 13 MCP tools with audit/budget/undo/risk-tier harness

    Python

  6. TrueNAS-AIops TrueNAS-AIops Public

    Python

Repositories

Showing 7 of 7 repositories

Top languages

Loading…

Most used topics

Loading…