Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add automateyournetwork/netclaw --skill hardware-health-checkgit clone --depth 1 https://github.com/automateyournetwork/netclawWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/automateyournetwork/netclaw/hardware-health-check)<a href="https://agentmods.dev/skills/automateyournetwork/netclaw/hardware-health-check"><img src="https://agentmods.dev/badge/skills/automateyournetwork/netclaw/hardware-health-check/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/automateyournetwork/netclaw/hardware-health-check"><img src="https://agentmods.dev/badge/skills/automateyournetwork/netclaw/hardware-health-check.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00067 | $0.01428 |
| Opus 5 | $0.00034 | $0.00714 |
| Sonnet 5 | $0.00013 | $0.00286 |
| Haiku 4.5 | $0.00007 | $0.00143 |
Grade A, and why
hardware-health-check scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Hardware Health Check (out-of-band, read-only)
MCP Server
- Server:
redfish-mcp(NetClaw-authored, spec 094) - Tools:
redfish_status,redfish_systems,redfish_thermal_power,redfish_managers,redfish_firmware,redfish_logs - Requires:
REDFISH_URL,REDFISH_USERNAME,REDFISH_PASSWORD - Read-only. No power control exists here — see below.
The one distinction this skill is for
NetClaw could not previously tell "the box is dead" from "the network to the box is dead." A BMC can, because it answers when the operating system cannot. But the distinction is symmetric, and each direction is a different wrong answer:
| Reading | What it establishes | What you must NOT say |
|---|---|---|
| BMC unreachable | nothing about the host | "the host is down" |
BMC reachable, Off |
the host is powered off — a fact | — |
BMC reachable, On |
the host has power | "the host is healthy / the OS is up" |
BMC reachable, health Critical |
a hardware fault is asserted | anything about the OS |
The BMC has its own NIC, its own network path and its own credentials, all separate from the host's. A BMC timeout is a statement about the BMC path, not about the server. Reporting "host down" from a BMC timeout is precisely the mistake out-of-band access exists to prevent.
Every response carries a verdict saying which of the rows above applies. redfish-mcp will
not emit a host claim without one — the tool refuses, so the qualifier cannot be dropped.
An auth rejection means the BMC is alive. HTTP 401/403 proves it answered; that is a credential problem, not a dead box, and the tool says so explicitly.
Workflow: is the box dead, or is it the network?
redfish_status— does the BMC answer at all?- If unreachable: stop and report exactly that. You have learned nothing about the host. Say which was tested (the BMC path) and what remains unknown (everything about the host).
- If reachable:
redfish_systems—PowerStateandStatus.Health - Read the
verdict, and phrase the answer in its terms.Offis a conclusion;Onis not. - If
Onbut the service is unreachable in band, the finding is "powered on, not serving" — which points at the OS, the application or the data network, not at the hardware.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 101 lines · 67 tokens per session scan A a96e7223a2bd
hardware-health-check is a skill published in the GitHub repository automateyournetwork/netclaw (655 stars, last pushed 5d ago), licensed Apache-2.0. It adds 67 tokens to every session and 1,428 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
ipmi-bmc-manager
IPMI / BMC management atomic skill. Controls server power, opens Serial-over-LAN consoles, reads sensors and event logs, and configures the BMC LAN on out-of-band controllers (Dell iDRAC, HPE iLO, Supermicro, generic IPMI 2.0) via ipmitool — in-band or over LAN. Use for remote power control and for recovering a node…
rtl-bug-repro
Reproduce an RTL bug with a minimal TB and isolate root cause via waveform diff — 'reproduce this bug', 'minimal repro TB', regression failure isolation.
lint-tool-profiles
Internal reference: lint tool profiles (agent-loaded; do not invoke).
boot-doctor-mcp
Deterministic PC no-boot / no-POST / no-display troubleshooting — real BIOS beep-code lookup and a ranked next-steps engine, no LLM guessing at hardware causes. Use this whenever the user describes a computer that won't turn on, won't POST, has no display output, is beeping, or asks for hardware troubleshooting steps…
gke-ai-troubleshooting-tpu-dynamic-slices-monitoring
Monitors, troubleshoots, and manages GKE TPU Dynamic Slices custom resources. Use when checking TPU slice lifecycle states, troubleshooting slice provisioning failures, validating single-slice or multi-slice (JobSet) workload manifests, or safely patching stuck finalizers and disabling the slice controller. Don't use…
gke-ai-troubleshooting-tpu-vbar-oom
Diagnoses and prevents vbarcontrolagent segfaults, out-of-memory (OOM) errors, and TPU device initialization failures on TPU v6e nodes in GKE caused by race conditions during TPU device resets or high-frequency metrics polling. Use when troubleshooting vbarcontrolagent crashes, memory cgroup OOMs in serial console…