Unity Cleaner Agent Skill

Clean up and audit a Unity project — find unused assets, content-hash duplicates, missing references, and empty folders. Use when reducing project bloat, hunting duplicate or orphaned assets, or auditing before a release, even if the user just says "清理项目" or "瘦身". 清理与审计 Unity 工程(查找未使用资源、内容哈希重复项、丢失引用、空文件夹);当用户要给项目瘦身、排查重复或孤立资源、或发布前审计时使用。

3k tokens

context cost

the whole folder, loaded on every use

files

instructions only

copies elsewhere

how many repositories repackaged it

1536

stars on the repo

on the repository, not the skill itself

Install

one command, takes just this skill from the repository

npx skills add https://github.com/Besty0728/Unity-Skills --skill unity-cleaner

The instruction itself

16 sections, as written by the author

Unity Cleaner Skills

> Safety: cleaner_delete_assets uses a two-step confirmToken handshake. Call without confirmToken to preview; call again with the returned token (5-minute TTL) to actually delete. There is no dryRun parameter.

Guardrails

Operating Mode (v1.9 three-tier):

Approval (default): all analyze/query skills (cleaner_find_unused_assets, cleaner_find_duplicates, cleaner_find_missing_references, cleaner_get_asset_usage, cleaner_find_empty_folders, cleaner_find_large_assets, cleaner_get_dependency_tree) are SemiAuto — run directly.
Auto / Bypass: SemiAuto and FullAuto run directly.
Auto-forbidden in this module: cleaner_delete_assets, cleaner_delete_empty_folders (both carry SkillOperation.Delete). In Approval/Auto these return MODE_FORBIDDEN — they are reachable only under Bypass mode or via a user-managed Allowlist entry; the grant flow does not unlock them.
cleaner_fix_missing_scripts is an Execute | Modify operation (it replaces missing component references with null in place, no asset deletion), so it does not trigger the NeverInSemi gate. In Auto / Bypass it runs directly; in Approval it still requires the standard grant handshake before execution.
cleaner_delete_assets additionally uses a two-step confirmToken handshake even when the mode gate allows it — preview first (no token), then confirm with the returned token (5-minute TTL).

DO NOT (common hallucinations):

cleaner_delete / cleaner_remove do not exist → cleaner skills only find/report; use asset_delete to actually remove
cleaner_fix does not exist → use cleaner_fix_missing_scripts specifically for missing script references
cleaner_scan / cleaner_find_unused do not exist → use specific skills: cleaner_find_unused_assets, cleaner_find_duplicates, cleaner_find_missing_references, cleaner_find_empty_folders, cleaner_find_large_assets

Routing:

To delete found assets → use asset module's asset_delete / asset_delete_batch
For project validation → use validation module

Skills Overview

| Skill | Description |

|-------|-------------|

| cleaner_find_unused_assets | Find assets not referenced by others |

| cleaner_find_duplicates | Find duplicate files by content hash |

| cleaner_find_missing_references | Find missing scripts/asset references |

| cleaner_delete_assets | Delete assets via two-step confirmToken (preview → confirm) |

| cleaner_get_asset_usage | Find what references a specific asset |

| cleaner_find_empty_folders | Find empty folders in the project |

| cleaner_find_large_assets | Find largest assets by file size |

| cleaner_delete_empty_folders | Delete all empty folders |

| cleaner_fix_missing_scripts | Remove missing script components from GameObjects |

| cleaner_get_dependency_tree | Get dependency tree for an asset |

Skills

cleaner_find_unused_assets

Find potentially unused assets of a specific type.

|-----------|------|----------|---------|-------------|

| limit | int | No | 100 | Max results |

Returns: {success, assetType, potentiallyUnusedCount, assets: [{path, name, type, sizeBytes}]}

# Find unused materials
result = call_skill("cleaner_find_unused_assets", assetType="Material")

# Find unused textures in a specific folder
result = call_skill("cleaner_find_unused_assets", 
    assetType="Texture2D", searchPath="Assets/Textures")

cleaner_find_duplicates

Find duplicate files by MD5 hash.

|-----------|------|----------|---------|-------------|

| limit | int | No | 50 | Max groups |

Returns: {success, duplicateGroupCount, totalWastedBytes, totalWastedMB, groups: [{count, sizeBytes, wastedBytes, files}]}

# Find duplicate textures
result = call_skill("cleaner_find_duplicates", assetType="Texture2D")
print(f"Wasted space: {result['totalWastedMB']:.2f} MB")

cleaner_find_missing_references

Find components with missing scripts or null references.

|-----------|------|----------|---------|-------------|

Returns: {success, issueCount, missingScripts, missingReferences, issues: [{type, gameObject, path, ...}]}

# Scan for all missing references
result = call_skill("cleaner_find_missing_references")
print(f"Missing scripts: {result['missingScripts']}")
print(f"Missing references: {result['missingReferences']}")

cleaner_delete_assets

Delete specified assets with two-step confirmation.

> ⚠️ Safety First: Deletion requires TWO calls - first preview, then confirm.

Step 1 - Preview (no confirmToken):

|-----------|------|----------|-------------|

Returns: {action: "preview", confirmToken, assetsToDelete, message}

Step 2 - Confirm (with confirmToken):

|-----------|------|----------|-------------|

Returns: {action: "deleted", deletedCount, results}

# Step 1: Preview what will be deleted
preview = call_skill("cleaner_delete_assets", 
    paths=["Assets/Unused/mat1.mat", "Assets/Unused/mat2.mat"])
print(preview['message'])  # Shows: "⚠️ PREVIEW ONLY - 2 assets will be deleted..."
print(preview['assetsToDelete'])  # Full list with sizes

# Step 2: Confirm deletion using token (expires in 5 minutes)
if input("Proceed? (y/n): ") == 'y':
    result = call_skill("cleaner_delete_assets", 
        confirmToken=preview['confirmToken'])
    print(result['message'])  # "Successfully deleted 2 assets"

cleaner_get_asset_usage

Find what objects reference a specific asset.

|-----------|------|----------|-------------|

| limit | int | No | Max results (default 50) |

Returns: {success, asset, usedByCount, usedBy: [{path, name, type}]}

# Check what uses a texture
result = call_skill("cleaner_get_asset_usage",
    assetPath="Assets/Textures/player.png")

`cleaner_find_empty_folders`

Find empty folders in the project.

|-----------|------|----------|---------|-------------|

Returns: { success, count, folders }

# Find empty folders
result = call_skill("cleaner_find_empty_folders")
print(f"Found {result['count']} empty folders")

`cleaner_find_large_assets`

Find largest assets by file size.

|-----------|------|----------|---------|-------------|

| limit | int | No | 20 | Max results |

Returns: { success, count, assets: [{ path, sizeBytes, sizeMB }] }

# Find top 10 largest assets over 1 MB
result = call_skill("cleaner_find_large_assets", limit=10, minSizeBytes=1048576)
for a in result['assets']:
    print(f"{a['sizeMB']:.2f} MB - {a['path']}")

`cleaner_delete_empty_folders`

Delete all empty folders.

|-----------|------|----------|---------|-------------|

Returns: { success, deleted, total }

# Delete all empty folders
result = call_skill("cleaner_delete_empty_folders")
print(f"Deleted {result['deleted']} of {result['total']} empty folders")

`cleaner_fix_missing_scripts`

Remove missing script components from GameObjects.

|-----------|------|----------|---------|-------------|

Returns: { success, removedComponents }

# Remove all missing script components
result = call_skill("cleaner_fix_missing_scripts")
print(f"Removed {result['removedComponents']} missing script components")

`cleaner_get_dependency_tree`

Get dependency tree for an asset.

|-----------|------|----------|---------|-------------|

Returns: { success, assetPath, dependencyCount, dependencies: [{ path, type }] }

# Get full dependency tree for a prefab
result = call_skill("cleaner_get_dependency_tree",
    assetPath="Assets/Prefabs/Player.prefab")
print(f"Dependencies: {result['dependencyCount']}")

Example Workflow: Clean Project

import unity_skills

# 1. Find all missing references and fix them
missing = unity_skills.call_skill("cleaner_find_missing_references")
for issue in missing['issues']:
    if issue['type'] == 'MissingScript':
        print(f"⚠️ Missing script on: {issue['path']}")

# 2. Find duplicate textures
dupes = unity_skills.call_skill("cleaner_find_duplicates", assetType="Texture2D")
if dupes['totalWastedMB'] > 10:
    print(f"🗑️ {dupes['totalWastedMB']:.1f} MB wasted on duplicates")

# 3. Find unused materials
unused = unity_skills.call_skill("cleaner_find_unused_assets", assetType="Material")
print(f"📦 {unused['potentiallyUnusedCount']} potentially unused materials")

# 4. Preview cleanup
paths_to_delete = [a['path'] for a in unused['assets'][:5]]
preview = unity_skills.call_skill("cleaner_delete_assets", 
    paths=paths_to_delete)
print(f"Would free: {preview['totalMB']:.2f} MB")
# To actually delete, call again with preview['confirmToken'] within 5 minutes.

Exact Signatures

Exact names, parameters, defaults, and returns are defined by GET /skills/schema or unity_skills.get_skill_schema(), not by this file.

Other skills for the same job

different authors, same section of the catalogue

Doc Coauthoring

by anthropics

vendor ×10

Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.

Intelligently organizes your files and folders across your computer by understanding context, finding duplicates, suggesting better structures, and automating cleanup tasks. Reduces cognitive load and keeps your digital workspace tidy without manual effort.

3k tokens

Domain Name Brainstormer

by frostant

×8

Generates creative domain name ideas for your project and checks availability across multiple TLDs (.com, .io, .dev, .ai, etc.). Saves hours of brainstorming and manual checking.

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

Implements Manus-style file-based planning for complex tasks. Creates task_plan.md, findings.md, and progress.md. Use when starting complex multi-step tasks, research projects, or any task requiring >5 tool calls.

9k tokens scripts

Scientific Brainstorming

by christophacham

×3

Creative research ideation and exploration. Use for open-ended brainstorming sessions, exploring interdisciplinary connections, challenging assumptions, or identifying research gaps. Best for early-stage research planning when you do not have specific observations yet. For formulating testable hypotheses from data use hypothesis-generation.

5k tokens

GitHub Project Management

by ComeOnOliver

×3

Comprehensive GitHub project management with swarm-coordinated issue tracking, project board automation, and sprint planning

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".

3k tokens

How to use it

Copy the folder

Take besty0728/unity-cleaner from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.