# Bare Metal

> Physical server provisioning, colocation, and dedicated hardware for performance-critical workloads

- Skill: `majiayu000/bare-metal` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds add majiayu000/bare-metal`
- Raw SKILL.md: https://api.skillmd.com/api/skills/majiayu000/bare-metal/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: majiayu000 (https://skillmd.com/u/majiayu000)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/majiayu000/bare-metal

---


# Skill: bare-metal

## What I do

I guide physical server provisioning, colocation management, and dedicated hardware deployment for high-performance computing, GPU workloads, and scenarios requiring full hardware control.

## When to use me

- High-performance computing and GPU-intensive workloads
- Full hardware control for optimisation and tuning
- Compliance requirements mandating physical isolation
- Cost optimisation at scale (large stable workloads)
- Latency-sensitive applications requiring bare metal performance

## Core principles

1. Automation over manual provisioning (PXE boot, Ansible)
2. Configuration management for reproducible deployments
3. Monitor everything: hardware health, temperature, disk SMART
4. Hardware redundancy (RAID, dual PSU, spare components)
5. Disaster recovery with offsite backups and runbooks

## Decision triggers

- Load with `devops` for provisioning automation
- Load with `scripter` for hardware management scripts
- Load with `automation` for deployment orchestration
- Load with `monitoring` for hardware health tracking
- For provisioning patterns, refer to Obsidian vault

## KB Reference

`~/vaults/baphled/3. Resources/Knowledge Base/AI Development System/Skills/DevOps-Operations/Bare Metal.md`