88lin/computer-repair-skill

让 Codex、Claude Code 等 AI Agent 成为你的跨平台电脑维修助手：64 个按需加载的 Playbook，覆盖诊断、清理、性能、网络与安全维护——C 盘爆满、电脑卡顿、流氓软件、应用迁移、数据恢复都能直接问；先取证、再计划，确认后修改。

📦 What is Computer Repair Skill?

Computer Repair Skill is a plug‑in (called a Skill) that can be added to AI‑powered coding assistants such as Codex, Claude Code, OpenClaw, Cursor, etc. Once installed, the Skill gives the Agent a library of 64 ready‑made “Playbooks” that describe step‑by‑step, evidence‑first procedures for diagnosing and fixing common computer problems on Windows, macOS and Linux.

The core idea is:

  1. Collect evidence first – read logs, system state, hardware info, disk usage, network config, etc.
  2. Explain possible causes – present a ranked list without changing anything.
  3. Propose a minimal‑impact plan – show the exact commands, their impact, rollback steps and verification.
  4. Execute only after explicit user confirmation.

The Skill is pure data (Markdown/YAML) plus a few helper scripts; it does not ship a GUI or a separate AI model. The host Agent does the heavy lifting (running commands, prompting the user, keeping a log).

🛠️ How to get it running

Method Who it’s for Quick steps
One‑sentence chat New users who already have an Agent that can fetch Git repos Send the Agent the sentence in the README (it tells the Agent to clone the repo, review SKILL.md, and install the Skill).
Skills CLI Users comfortable with Node.js/npm npx skills add 88lin/computer-repair-skill (add --global and `--agent codex
Repository installer Users who want full control over the install location Clone the repo, then run install.ps1 (Windows) or install.sh (macOS/Linux) with --target (codex, claude, agents or custom).
Manual copy Environments where command‑line tools are unavailable Copy the whole skills/computer-repair-skill folder into the Agent’s skills directory.

All installers verify the target path, back up any existing version, and refuse to overwrite symbolic links or junctions. They also lock the install process to avoid concurrent runs.

🎯 What it can do (selected examples)

System health & storage (20 Playbooks)

  • Full health check – baseline report for hand‑over or pre‑repair documentation.
  • Windows / macOS / Linux performance profiling (CPU, memory, I/O, container limits).
  • Disk‑space audit – list biggest folders, safe‑to‑delete caches, and migration plans.
  • Application migration – copy data, verify integrity, create a NTFS junction, and keep a rollback plan.
  • New‑machine acceptance – hardware inventory, serial numbers, first‑time setup checklist.

Network & identity (9 Playbooks)

  • Wi‑Fi / Ethernet diagnostics – layer‑by‑layer check from link up to DNS.
  • VPN troubleshooting – split‑tunnel, DNS leak, and endpoint reachability.
  • Email connectivity tests – SMTP/IMAP/POP3 port and TLS checks without sending real mail.

Apps, updates & printers (8 Playbooks)

  • Uninstall residue cleanup – registry backup, file‑system audit, recycle‑bin safety net.
  • Outlook sync issues – isolate network vs. profile problems.
  • Printer queue recovery – CUPS on macOS, Spooler on Windows.

Windows‑specific repair (7 Playbooks)

  • Driver audit – identify hardware IDs, locate official sources, plan rollback.
  • Boot‑failure triage – differentiate missing boot device, UEFI/Legacy mis‑config, or corrupted BCD.
  • WinRE workflow – SFC/DISM, safe‑mode boot, restore points.
  • BitLocker pre‑check before entering PE – status, TPM health, recovery‑key location.

Data recovery (1 Playbook)

  • Read‑only imaging, scan, and report before any write operation.

Security & credentials (5 Playbooks)

  • Endpoint security audit – AV, firewall, suspicious activity logs.
  • Browser policy review – extensions, password storage, AI‑related telemetry.
  • Credential inventory – SSH keys, cloud tokens, with no secret values printed.

Development environment (3 Playbooks)

  • Homebrew install & health check on macOS.
  • CUDA driver & toolkit verification on Linux.
  • SSH‑key generation and GitHub configuration.

OpenClaw integration (11 Playbooks)

  • Install, configure, and troubleshoot the OpenClaw LLM gateway (Node.js version, model selection, channel plugins for Feishu, Telegram, WhatsApp, etc.).
  • Safe uninstall that lists services, config files and data directories before deletion.

All actions follow a read‑first, confirm‑then‑write policy and include explicit rollback instructions.

🔐 Safety & privacy guarantees

  • Read‑only first – every Playbook starts with evidence collection; no file or registry changes happen until the user explicitly approves.
  • No secret leakage – passwords, recovery keys, private tokens are never echoed; only hashes or masked values are shown.
  • No dangerous defaults – the Skill never disables Defender, SmartScreen, UAC, Windows Update, or firewall.
  • Offline‑compatible – the Skill assumes the host Agent may lose network (e.g., when booted into WinRE). All steps are documented so a human can continue without the Agent.
  • Audit‑ready – every command is logged, and the Skill can run a Git‑storage audit script (scripts/git_storage_audit.py) that outputs a JSON inventory of local repositories.

📚 Who should use it?

  • IT support technicians who want a reproducible, evidence‑driven checklist that can be run by an AI assistant.
  • Power users who frequently clean disks, migrate apps, or troubleshoot network issues across Windows/macOS/Linux.
  • Developers of AI agents looking for a ready‑made Skill package that demonstrates safe system‑modifying workflows.
  • Organizations that need a documented, auditable repair process that can be invoked from a conversational interface.

🗂️ Repository layout (quick glance)

skills/computer-repair-skill/   # core Skill package
│   SKILL.md          # routing table & safety workflow
│   agents/openai.yaml # metadata for Codex display
│   references/       # 64 Playbook markdown files + index
│   scripts/          # install scripts, git‑audit helper
│   tests/validate_skill.py  # CI validator (no external deps)
assets/                # cover illustration
docs/                  # static site (GitHub Pages) – bilingual UI
tools/                 # data‑extraction & site‑build helpers
LICENSE, NOTICE, SECURITY.md, CONTRIBUTING.md

✅ Bottom line

Computer Repair Skill is a genuine, open‑source project that equips AI coding assistants with a comprehensive, safety‑first toolbox for real‑world computer diagnostics and repair. It is not a tutorial list or a trivial demo; it ships a structured set of 64 Playbooks, installation helpers, CI validation, and a clear security model, making it a useful asset for anyone who wants an AI‑augmented, auditable approach to PC maintenance.

Related

  • Project
    cisco-ai-defense/skill-scanner

    Skill Scanner is an open‑source security scanner for AI‑agent skill packages (e.g., OpenAI Codex, Cursor, Claude). It combines signature‑based YARA/YAML rules, AST‑based data‑flow analysis, optional LLM‑as‑a‑judge, and a CEL decision layer to detect prompt‑injection, data‑exfiltration, and malicious code patterns. Install via pip, run from the CLI or Python SDK, and integrate with CI/CD via GitHub Actions or a pre‑commit hook. Findings are reported with severity levels; a clean scan is best‑effort only and does not guarantee safety.

  • Project
    MoizIbnYousaf/ai-agent-skills

    A genuine npm‑based CLI/TUI that ships a curated library of AI‑agent skills, lets you build and manage your own skill collections, and integrates with many LLM‑agents via the Agent Skills spec.

  • Project
    GPTomics/bioSkills

    bioSkills is a library of ready‑made code templates (“skills”) that teach LLM coding agents how to perform bioinformatics analyses—from sequence I/O to single‑cell, variant calling, structural biology, chemoinformatics, and workflow management. Installable via thin scripts for Claude Code, Codex, Gemini/Antigravity, OpenCode, and OpenClaw, the skills provide best‑practice command‑line flags, example scripts and usage guides, improving the correctness of AI‑generated bio‑informatics code.

  • Project
    NVIDIA/SkillSpector

    SkillSpector is an open‑source security scanner for AI‑agent “skills”. It statically checks code bundles for 71 known vulnerability patterns (prompt injection, data exfiltration, privilege escalation, etc.) and can optionally run an LLM‑based semantic analysis. It accepts directories, zip files, Git URLs, or single SKILL.md files, produces terminal, JSON, Markdown, or SARIF reports, and supports baseline suppression to hide known false positives. Install via `uv tool install` or Docker, and configure many LLM providers (OpenAI, Anthropic, Bedrock, NVIDIA Build, Gemini, Ollama, Azure, etc.) for the optional second stage. Designed for developers and security teams to vet skills before publishing them to NVIDIA’s Verified Skills catalog or internal registries.

  • Project
    amd/skills

    AMD Skills is a catalog of reusable, agent‑compatible “skills” that encode AMD‑specific AI workflow knowledge (ROCm diagnostics, Quark quantisation, LLM serving on Instinct/EPYC, etc.). Install via the `skills` CLI (`npx skills add amd/skills`) or manually, then invoke the skills through coding assistants like Cursor, Claude Code, Codex or Gemini. Each skill bundles a `SKILL.md` description, optional scripts and references, enabling agents to automatically perform setup, optimisation and debugging steps for AMD hardware.