Workstation Maintenance logo

Workstation Maintenance

Community
appleweiping
workstation-maintenance

Inventory and tidy local Windows drives without breaking anything. Use when asked to clean up or organize drives or folders, plan file moves in small approval batches, move cluttered top-level folders of a drive behind junctions, or undo an earlier move. The folders to protect come from the user's config.json; nothing moves without an approved batch ID and a rollback manifest.

Overview

Publisherappleweiping
RepositoryAGENT_RESOURCE
Skill nameworkstation-maintenance
Stars
78
Forks
10
Bundled files
15
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 15 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by appleweiping on GitHub. Read the source before you install it.

Installation

Install the Workstation Maintenance AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/appleweiping/AGENT_RESOURCE/tree/master/skills/vipin/workstation-maintenance
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/appleweiping/AGENT_RESOURCE.git /tmp/AGENT_RESOURCE
mkdir -p .claude/skills
cp -r /tmp/AGENT_RESOURCE/skills/vipin/workstation-maintenance .claude/skills/workstation-maintenance
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Workstation Maintenance in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Workstation Maintenance on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Workstation Maintenance is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Workstation Maintenance

Use this skill to inventory Windows drives, propose file moves in small batches, move only the batches the user approves, and roll them back on request. A second mode moves cluttered top-level folders of a drive into <drive>\_Organized\<Bucket>\_RootDirs\ and leaves a junction at the old path, so existing absolute paths keep working.

Setup: config.json

Everything specific to the machine lives in config.json next to this file. It is git-ignored. If it is missing, copy config.example.json, ask the user which folders must never move, and fill it in before any live run. A live inventory and the drive-root scripts refuse to run without a config. -ConfigPath or $env:WORKSTATION_MAINTENANCE_CONFIG can point at a config elsewhere. Paths may use environment variables such as %USERPROFILE%.

KeyMeaning
protected_rootsNever scanned, never moved, never a destination.
infrastructure_rootsAgent and tool installs. Recorded, never moved.
project_rootsActive project folders. Recorded, never moved.
inventory_rootsWhat to scan when -Roots is not given.
organized_rootWhere file-level moves go when -TargetRoot is not given.
coursework_patternsExtra filename regexes treated as coursework (kept in place).
drive_root_rules{ "pattern", "category" } pairs: regexes on top-level folder names mapped to a drive-root bucket.

Agent folders in the user profile (.codex, .claude, .openhands, opencode state) are always treated as infrastructure.

Hard boundaries

  • Never move, delete, or open anything under a protected_roots entry, including when it is reached through a junction, symlink, relative path, or different casing.
  • The Windows folder, Program Files, ProgramData, AppData, package caches, browser profiles, auth/session/db/log folders, git worktrees and reparse points are never moved.
  • infrastructure_roots and project_roots stay in place unless the user asks for a migration of that one folder.
  • Never delete files. At most, write a list of deletion candidates for the user.
  • Classify by path and filename only. Do not open documents.
  • No physical move without an inventory manifest, a move plan, a batch ID the user approved, and rollback data.

Workflow

  1. Preflight
    • Check that config.json exists and protects what the user cares about.
    • Note uncommitted changes in any repos under the roots and leave them alone.
  2. Inventory
    • Run scripts/New-WorkstationInventory.ps1. Keep its JSON/Markdown output in a local folder that is not committed; it contains private filenames.
    • Confirm there are zero entries under protected roots and zero move-eligible reparse points, directories, or git worktrees.
  3. Plan
    • Run scripts/New-MovePlan.ps1 on the inventory manifest.
    • Executable batches include only low-risk files older than 30 days; newer ones are deferred so normal use is not disrupted.
    • Batches are grouped by file type and capped at 100 items, so the user approves small, understandable units. Duplicate destinations get a stable item-ID suffix.
    • For drive-root organization, run scripts/New-DriveRootOrganizationPlan.ps1 -DriveRoot <drive> instead. It classifies the immediate children of the drive only.
  4. Approval gate
    • Stop before moving anything.
    • scripts/New-ApprovalPacket.ps1 writes a summary of batch IDs, counts and sizes, without filenames. scripts/Test-MovePlanBatches.ps1 preflights every batch.
    • Ask the user to approve batch IDs. If they approve all currently passing low-risk batches at once, run them without asking again for each one, stopping at the first failure.
    • Only then run scripts/Invoke-ApprovedMoveBatch.ps1 -Approved (or Invoke-DriveRootOrganizationPlan.ps1 -Approved).
  5. Rollback and validation
    • Every applied batch writes an applied manifest with rollback fields.
    • Undo with scripts/Invoke-RollbackBatch.ps1, or Invoke-DriveRootOrganizationPlan.ps1 -RollbackManifestPath <applied.json>.
    • Run scripts/Test-WorkstationMaintenance.ps1 after changing any script.
  6. Report
    • Summarize what moved by bucket, count and size. Do not paste private filenames into chat logs, notes, or commits.

Categories

Protected-NoMove, AgentInfrastructure, ActiveProject, CourseworkArchive, PersonalSensitive, MediaAssets, Downloads, TempCache, VendorSystemToolchain, UnknownReview. Drive-root buckets: Downloads, Temp-Review, Media, Coursework, Documents-Private, Games, Tools-Review.

Commands

Run from this folder. powershell -NoProfile -ExecutionPolicy Bypass -File <script> ... works the same if the execution policy blocks .\scripts\....

powershell
$out = Join-Path $env:TEMP "workstation-maintenance"
.\scripts\New-WorkstationInventory.ps1 -OutputDir $out
.\scripts\New-MovePlan.ps1 -ManifestPath "<inventory.json>"
.\scripts\Test-MovePlanBatches.ps1 -MovePlanPath "<move-plan.json>"
.\scripts\New-ApprovalPacket.ps1 -MovePlanPath "<move-plan.json>"

New-MovePlan.ps1 defaults to -MinimumAgeDays 30 -MaxItemsPerBatch 100. Lower the age gate only when the user accepts the risk of moving recent files.

After the user names a batch ID:

powershell
.\scripts\Invoke-ApprovedMoveBatch.ps1 -MovePlanPath "<move-plan.json>" -BatchId "batch-downloads-archives-old" -PreflightOnly
.\scripts\Invoke-ApprovedMoveBatch.ps1 -MovePlanPath "<move-plan.json>" -BatchId "batch-downloads-archives-old" -Approved
.\scripts\Invoke-RollbackBatch.ps1 -AppliedManifestPath "<applied-batch.json>"

Drive-root organization (<drive> is a drive root such as the one the user named):

powershell
.\scripts\New-DriveRootOrganizationPlan.ps1 -DriveRoot "<drive>" -OutputDir $out
.\scripts\Invoke-DriveRootOrganizationPlan.ps1 -PlanPath "<drive-root-plan.json>" -PreflightOnly
.\scripts\Invoke-DriveRootOrganizationPlan.ps1 -PlanPath "<drive-root-plan.json>" -Approved -SkipIds @("dr_0016")

Folders that fail because of ACLs or open handles stay where they are: pass their IDs through -SkipIds, report them, and retry after the lock is gone. Never leave a folder half moved.

References

  • references/safety-boundaries.md
  • references/batch-move-workflow.md
  • references/manifest-schema.md

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Workstation Maintenance AI skill do?

Inventory and tidy local Windows drives without breaking anything. Use when asked to clean up or organize drives or folders, plan file moves in small approval batches, move cluttered top-level folders of a drive behind junctions, or undo an earlier move. The folders to protect come from the user's config.json; nothing moves without an approved batch ID and a rollback manifest.

Why use Workstation Maintenance on TypingMind?

Because you install it once and use it with any model. Workstation Maintenance is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Workstation Maintenance in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/appleweiping/AGENT_RESOURCE/tree/master/skills/vipin/workstation-maintenance. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Workstation Maintenance?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Workstation Maintenance?

As many as you like. As long as a model supports skills, you can use Workstation Maintenance with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Workstation Maintenance AI skill free?

Yes. It is published on GitHub by appleweiping under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇