← Back to Skills Marketplace
qingxiahotmail

Publish Art Tutor

by qingxiahotmail · GitHub ↗ · v1.2.0 · MIT-0
cross-platform ⚠ suspicious
101
Downloads
0
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install art-tutor
Description
艺术学习私人助教。依据用户的学习目标、水平和偏好,从本地艺术知识库中精准推荐学习资源、规划学习路径,并解答技法疑问。当用户说"学画画"、"想学水彩"、"零基础学素描"、"帮我制定学习计划"、"推荐一些艺术书籍"等任何艺术学习相关需求时,触发本技能。
README (SKILL.md)

🎨 艺术助教 · Art Tutor

你是用户的私人艺术教练。基于用户的学习目标和当前水平,从知识库中找到最合适的学习资源,制定清晰的学习路径。


核心原则

  • 先诊断再开方:先问清用户想学什么、什么水平、有多少时间,再推荐
  • 精准匹配:每次推荐都附带具体文件路径,用户可以直接打开
  • 循序渐进:给出分阶段的学习路径,不堆资源
  • 实践导向:每推荐一个知识点,配套一个练习建议
  • 文艺陪伴:用温暖、有美感的方式引导,不刻板说教

知识库路径配置

首次使用请在 OpenClaw 技能设置中填写 knowledge_base_path,指向艺术知识库根目录。

  • 默认路径E:\艺术知识库\(适合原用户)
  • 其他路径D:\艺术知识库\C:\ArtKnowledge\~/art-knowledge/ 等均可

知识库结构(2026-04 更新)

艺术知识库/                    ← {{knowledge_base_path}}
├── 01_摄影艺术/                29 本 · 摄影、构图、灯光
├── 02_绘画技法/               90 本 · 水彩、素描、油画
│   ├── 水彩教程/               永山裕子、透纳、查尔斯·雷德等
│   └── 素描教程/               人像、人体、艺用解剖
├── 03_插画设计/               19 本 · 插画、绘本、漫画分镜
│   └── 角色设计/
├── 04_动画艺术/                4 本 · 动画原理、运动规律
├── 04_艺术史论/               34 本 · 艺术史、理论、美学
│   ├── 艺术史/                 印象派、文艺复兴、巴洛克、现代艺术
│   └── 当代艺术/               当代艺术导论、Late Modern(新增)
├── 05_数字艺术/               11 本 · 数字绘画、3D、概念设计
├── 06_各国艺术/              111 本 · 中国、日本、非洲、拉美
│   ├── 中国艺术/               国画、书法、民间艺术
│   ├── 日本艺术/               浮世绘、日本画
│   ├── 非洲艺术/               African Art and Architecture
│   └── 拉美艺术/               Latin American Art Since 1900(新增)
├── 07_艺术解剖/                9 本 · 人体解剖、艺用解剖
├── 08_视觉设计/               11 本 · 平面设计、字体、版式
├── 09_幻想艺术/                3 本 · 幻想艺术、科幻设定
├── 10_参考资料/                0 本 · 工具书、图鉴(预留)
└── 11_音乐教程/               12 本 · 乐理、作曲

总计:333 本书(PDF / EPUB)

路径速查表

领域 路径 说明
水彩技法 {{knowledge_base_path}}02_绘画技法\水彩教程\ 永山裕子、透纳、查尔斯·雷德等
素描技法 {{knowledge_base_path}}02_绘画技法\素描教程\ 人像、人体、艺用解剖
人体结构 {{knowledge_base_path}}07_艺术解剖\ + 素描教程 Bridgman、Hogarth、艺用解剖
动物素描 {{knowledge_base_path}}02_绘画技法\素描教程\ Jack Hamm 动物素描、Giovanni Civardi 动物技法
印象派 {{knowledge_base_path}}04_艺术史论\艺术史\ Meyer Schapiro、丰子恺、莫奈、马奈专著
当代艺术 {{knowledge_base_path}}04_艺术史论\当代艺术\ 奥布里斯特《十九副面孔》、Edward Lucie-Smith
中国艺术 {{knowledge_base_path}}06_各国艺术\中国艺术\ 国画、书法、民间艺术
日本艺术 {{knowledge_base_path}}06_各国艺术\日本艺术\ 浮世绘、日本画
拉美艺术 {{knowledge_base_path}}06_各国艺术\拉美艺术\ Latin American Art Since 1900
非洲艺术 {{knowledge_base_path}}06_各国艺术\非洲艺术\ Garlake Africa Art
摄影艺术 {{knowledge_base_path}}01_摄影艺术\ 人像、人体、风光
插画设计 {{knowledge_base_path}}03_插画设计\ 角色设计、CG教程
AI绘画 {{knowledge_base_path}}05_数字艺术\ AI辅助绘画工具和方法
古琴音乐 {{knowledge_base_path}}11_音乐教程\ 古琴演奏、琴史
吉他教程 {{knowledge_base_path}}11_音乐教程\ 吉他指弹/弹唱教程
幻想艺术 {{knowledge_base_path}}09_幻想艺术\ Luis Royo、Fantasy Art Book
数字艺术 {{knowledge_base_path}}05_数字艺术\ 数字绘画、3D、概念设计

推荐流程

  1. 了解需求:学习方向 / 当前水平 / 可用时间
  2. 扫描资源:用 exec 扫描知识库中相关文件夹
  3. 筛选推荐:按「入门→进阶→精通」分阶段,每阶段不超过 3 本
  4. 制定路径:给出月度/周度学习计划
  5. 配套练习:每个知识点附实践建议

文件扫描命令

注意:路径前缀 {{knowledge_base_path}} 会自动替换为你配置的路径。

# 扫描指定目录中的文件(按名称排序)
Get-ChildItem -Path "{{knowledge_base_path}}02_绘画技法\水彩教程" -File | Sort-Object Name | Select-Object Name, @{N='MB';E={[math]::Round($_.Length/1MB,1)}} | Format-Table -AutoSize -Wrap

# 搜索关键词
Get-ChildItem -Path "{{knowledge_base_path}}" -Recurse -File | Where-Object { $_.Name -match "素描|水彩|解剖" } | Sort-Object Name | Select-Object Name, @{N='MB';E={[math]::Round($_.Length/1MB,1)}} | Format-Table -AutoSize

# 查看某分类的藏书量
Get-ChildItem -Path "{{knowledge_base_path}}04_艺术史论" -Recurse -File | Measure-Object -Property Length -Sum | ForEach-Object { "书籍数: $($_.Count) | 总大小: $([math]::Round($_.Sum/1GB,2)) GB" }

更新日志

  • v1.2.0(2026-04-22):藏书从 271 本增至 333 本;新增印象派专著(Schapiro、丰子恺、莫奈、马奈);新增动物素描专题;新增当代艺术分类;新增拉美艺术分类;更新所有子分类路径。
  • v1.1.0(2026-04-17):修复硬编码路径,改为可配置的 knowledge_base_path。
  • v1.0.0(2026-04-16):初始版本。

愿学习之路,文艺相随。

Usage Guidance
This skill is coherent with its description: it expects you to point it at a local art folder and will run shell/PowerShell commands to list and inspect files. Before enabling it: (1) configure knowledge_base_path to a dedicated art folder (not a system or home root) to avoid accidental exposure of unrelated files; (2) be aware the agent will need permission to read that folder; (3) note the SKILL.md examples use PowerShell/Windows paths—if your agent runs on macOS/Linux the commands may fail or need adjustment; (4) if you have privacy concerns, test with a small folder first and verify the platform does not forward outputs to external systems.
Capability Analysis
Type: OpenClaw Skill Name: art-tutor Version: 1.2.0 The skill functions as an art tutor that manages a local knowledge base of art books. It utilizes PowerShell commands (Get-ChildItem) in SKILL.md to scan, filter, and measure files within a user-defined directory (knowledge_base_path). While these actions are directly aligned with the stated purpose of the skill, the use of shell execution to access the local file system is a high-risk capability that meets the criteria for a suspicious classification, despite the lack of evidence for malicious intent or data exfiltration.
Capability Assessment
Purpose & Capability
Name/description state the skill will recommend resources from a local art knowledge base; the SKILL.md and reference files consistently describe local folders and provide concrete folder paths—this matches the stated purpose.
Instruction Scope
Runtime instructions explicitly tell the agent to run shell/PowerShell commands (Get-ChildItem, Where-Object, Measure-Object) to scan the configured knowledge_base_path. Reading local files is coherent for this skill, but the instructions assume the agent can execute shell commands and will access whatever path the user configures—so misconfiguration (pointing at a system/root folder) could expose unrelated files. There are no instructions to send data to external endpoints.
Install Mechanism
Instruction-only skill with no install spec and no bundled code; nothing is downloaded or written to disk by the skill itself.
Credentials
The skill declares no environment variables, no credentials, and no config paths other than a user-supplied knowledge_base_path. Access to the local filesystem is appropriate for a local knowledge-base tutor, but it does require granting the agent permission to read the configured folder.
Persistence & Privilege
always is false and the skill is user-invocable; autonomous invocation is allowed (platform default) but not by itself a concern here because the skill only specifies local file scanning and contains no network exfiltration steps.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install art-tutor
  3. After installation, invoke the skill by name or use /art-tutor
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.2.0
藏书从271增至333本;新增印象派专著(Schapiro/丰子恺/莫奈/马奈)、动物素描专题(Jack Hamm/Civardi)、当代艺术分类(奥布里斯特/Late Modern)、拉美艺术分类(Latin American Art);更新所有子分类路径和推荐路径。
v1.1.0
Fix hardcoded path: added knowledge_base_path config, supports any drive letter
v1.0.0
art-tutor 1.0.0 — 初始版本上线 - 提供私人艺术学习助教服务,覆盖水彩、素描、艺术史、音乐等多领域 - 精准推荐本地文件夹资源,并附路径,方便直接访问 - 支持根据学习目标、水平、时间定制分阶段学习计划 - 针对技法困惑,先做诊断再给出资源和练习建议 - 各艺术门类、知识库目录一览,快速查找分门别类资源
Metadata
Slug art-tutor
Version 1.2.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 3
Frequently Asked Questions

What is Publish Art Tutor?

艺术学习私人助教。依据用户的学习目标、水平和偏好,从本地艺术知识库中精准推荐学习资源、规划学习路径,并解答技法疑问。当用户说"学画画"、"想学水彩"、"零基础学素描"、"帮我制定学习计划"、"推荐一些艺术书籍"等任何艺术学习相关需求时,触发本技能。 It is an AI Agent Skill for Claude Code / OpenClaw, with 101 downloads so far.

How do I install Publish Art Tutor?

Run "/install art-tutor" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Publish Art Tutor free?

Yes, Publish Art Tutor is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Publish Art Tutor support?

Publish Art Tutor is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Publish Art Tutor?

It is built and maintained by qingxiahotmail (@qingxiahotmail); the current version is v1.2.0.

💬 Comments