hardware-health-check

Diagnose server hardware health and power state via read-only Redfish BMC queries.

650|182|Updated Feb 19, 2026
One-click install
npx skills add https://github.com/automateyournetwork/netclaw --skill hardware-health-check
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hardware-health-check
Source: https://github.com/automateyournetwork/netclaw/tree/main/workspace/skills/hardware-health-check
Command: npx skills add https://github.com/automateyournetwork/netclaw --skill hardware-health-check

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

When a server stops responding, you cannot tell whether the host itself is dead or only the network path to it is broken. This Skill uses out-of-band Redfish BMC access to distinguish a powered-off machine from a network failure, and to triage hardware faults, thermal issues, and firmware problems without touching the host OS.

Core Features & Use Cases

  • Power State Verification: Query the BMC to determine whether a host is powered off, powered on, or simply unreachable via the BMC path, with an explicit verdict on every response.
  • Hardware Fault Triage: Inspect component health rollups, thermal and power readings against critical thresholds, PSU status, and SEL event logs by severity.
  • Firmware Inventory: Review firmware versions across components and the BMC itself to identify known-bad levels.
  • Use Case: A production server stops responding to SSH. Use this Skill to check the BMC: if it reports PowerState Off, you have a confirmed fact; if the BMC itself is unreachable, you know only that the BMC path failed and must not claim the host is down.

Quick Start

Ask the agent to check whether the server at the configured Redfish BMC address is powered off or just unreachable, and report the verdict.

Frequently Asked Questions about hardware-health-check

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check if a server is powered off or just unreachable?

Query the server's BMC over Redfish using redfish_status and redfish_systems. If the BMC answers and reports PowerState Off, the host is confirmed powered off; if the BMC itself is unreachable, you have learned nothing about the host and must report only the BMC path failure.

What is Redfish BMC monitoring used for?

Redfish BMC monitoring provides out-of-band access to hardware state independent of the host OS. It exposes power state, component health rollups, thermal and power readings, firmware inventory, and SEL event logs through a standard REST API on controllers like iDRAC and iLO.

Can this skill remotely power cycle a server?

No. The skill is deliberately read-only and does not implement Redfish ComputerSystem.Reset or any power control. A required power cycle is an operator action through the BMC UI, gated by change control.

Why does an empty SEL log not mean the hardware is healthy?

SELs are ring buffers that get cleared during service, so zero entries means no recorded entries, not no faults. The tool surfaces this caveat in its gaps field rather than reporting a clean bill of health.

Does Redfish work with self-signed BMC certificates?

Yes, TLS verification is disabled by default because BMCs ship self-signed certificates, and every response discloses this. Set REDFISH_VERIFY_TLS=true to require a valid certificate on untrusted networks.