DeepSeek Harness Plugin Hub

发布与管理完整 Harness Profiles,发现适合你的插件。

探索

插件目录环境预设文档中心动态

社区

发布插件联系我们报告问题

相关链接

Plugin Hub GitHubDeepSeek Harness 官方项目系统状态隐私说明
© 2026 DeepSeek Harness Plugin HubPowered byPaxTech

独立、非官方社区项目,与 DeepSeek 官方无隶属、授权或背书关系。

Blender Mentor — DeepSeek Harness 插件(DSH Plugin)
DeepSeek Harness Plugin Hub
ProfilesPlugins分类动态文档登录管理 Profiles
ProfilesPlugins分类动态文档登录
← Plugins
B

blender-mentor

Blender Mentor

Blender 全栈导师 AI Skill:教学、0-3 评分标准评定、资产/动画验收(ESD/BIO)。

插件会安装到这里;不确定时保持 web。

npx -y @deepseek-ai/dsh plugin --profile web add github:Memory175/blender-mentor#c31b2b64570c084bbbec29b10260b457e5396243
README兼容性版本

兼容性与来源证明

Blender Mentor 以 blender-mentor 发布,当前版本为 5.4.2。Plugin Hub 会校验它的 manifest,并保存精确安装来源,便于复现安装结果。

DSH 兼容范围
*
运行环境
any
发布来源
github
Registry 更新时间
2026/10/19

版本

5.4.2stable
2026/10/19

相关插件

正在加载相关插件…

最新版
5.4.2
DSH
*
HMR
重启进程
Tree shaking
未声明可安全裁剪
解包体积
未提供
文件数
未提供
Surface
any
许可证
MIT
发布源
github
GitHub
★ 0
周下载
0
最近提交
2026/10/19
查看源码 ↗
README Badge

点击下方 Badge 复制 Markdown,粘贴到 README 即可。

这是你的 Plugin?认领权益 · 优先安全扫描

验证 package.json 声明的 GitHub 仓库,即可管理这个公开页面。认领后,Hub 会优先安排当前版本的安全扫描,并在通过后公开展示结果。

认领这个 Plugin →
报告问题

相关插件

继续浏览 productivity-workflow 分类下经过校验的插件。

Acp App@deepseek-ai/dsh-acp-appdsh ACP 配置文件包:基于 dsh-base 的仅限自动化的 JSON-RPC stdio 和进程生命周期管理Experimental Schedule Bundle@deepseek-ai/dsh-experimental-schedule-bundle为 Web 组合添加 Schedule、其任务页面和时间上下文的可选捆绑包Ab Ocrdsh-ab-ocr--- description: "外部 ocr 工具:逐页识别扫描的 PDF 或图像,仅在处理完最后一页后按顺序合并页面,并使用模型确定的标题级别保存为一个 Markdown 文档。" kind: "package-reference" ---Scriptor@linfengqaqtat/dsh-scriptorWebnovel Writer:DSH 长篇网文写作工作台

README

blender-mentor

Blender Mentor - AI Agent Skill banner

Blender Full-Stack Mentor V5.4.0 · AI Agent Skill (standard layout) Drop this repo into any Agent Skills–capable client (DSH / Claude Code / Claude.ai / Cursor) and let the agent teach, grade, review, and verify Blender work like a senior mentor + technical director.

English | 简体中文

What it is

SKILL.md is the resident operating manual (routing + D1–D4 delivery loop + machine-acceptance commands); references/ holds six on-demand persona documents; scripts/ ships executable acceptance tools; assets/ holds templates and rule data.

Core behavior:

  • Diagnose before acting — place the learner at an L0–L3 candidate level via the rubric, then choose explain / coach / accept / firefight / review mode.
  • Delivery loop D1→D4 — deliver → self-evaluate by sampling the artifact → P0/P1/P2 fix list → an explicit, option-based question.
  • Machine acceptance — asset gate, curve audit, render probe, ESD envelope checks, biomechanics checks.
  • No fabrication, no unauthorized actions, no sign-off substitution — acceptance evidence must be sampled from the artifact.

Repository layout

.
├── SKILL.md                 # Agent Skill entry point (required)
├── README.md                # English (this file)
├── README.zh-CN.md          # 简体中文
├── LICENSE                  # MIT
├── CHANGELOG.md
├── .gitignore
├── package.json             # DSH plugin manifest (dsh.bundle.patch)
├── cordis.patch.yml         # DSH Cordis patch (skill-filesystem provider)
├── lib/index.js             # DSH plugin adapter
├── skills/
│   └── blender-mentor/      # DSH bundled skill (generated)
├── .claude-plugin/
│   └── marketplace.json     # Claude Code plugin marketplace
├── .github/workflows/validate.yml
├── tools/
│   └── build_skill.py       # maintainer build tool (rebuild from the V5.4.0 six-file suite)
├── plugins/
│   └── blender-mentor/      # Claude Code plugin wrapper (generated)
├── dsh/
│   ├── install.ps1          # DSH one-line installer (Windows)
│   ├── install.sh           # DSH installer (Linux/macOS/WSL)
│   └── README.md
├── references/              # persona documents, loaded on demand
│   ├── main.md              # identity / boundaries / six-part template / 40 red lines
│   ├── rubric.md            # 0–3 scoring rubric / critical items / self-review template
│   ├── curriculum.md        # M0–M6 course package
│   ├── toolbox_text.md      # 130-misconception library (code stripped)
│   ├── module-esd.md        # ESD-001 energy-scaled secondary animation
│   └── module-biomechanics.md # BIO-001 biomechanics
├── scripts/                 # 10 executable scripts
│   ├── asset_gate.py        # asset gate (headless Blender)
│   ├── curve_audit.py       # curve audit
│   ├── render_probe.py      # render probe (//probe.png)
│   ├── envelope_check.py    # ESD envelope assertions
│   ├── expand_esd_spec.py   # expand ESD template to executable spec
│   ├── sample_esd_from_blend.py # sample a measured spec from a .blend
│   ├── biomech_check.py     # BIO-001 biomechanics acceptance
│   ├── persona_lint.py      # persona structure gate + deny-word scan
│   ├── load_router.py       # generate the load plan for a task
│   └── selftest.py          # toolchain self-test (strict three-state)
└── assets/
    ├── load_plan.json
    ├── esd_spec.template.json
    ├── esd_spec.example.json
    ├── esd_sample_config.template.json
    └── deny.example.txt

Installation

DSH (local)

One-line install on Windows (uses git clone; no dependence on raw.githubusercontent.com):

git clone --depth 1 https://github.com/Memory175/blender-mentor.git "$env:TEMP\blender-mentor-src"; powershell -NoProfile -ExecutionPolicy Bypass -File "$env:TEMP\blender-mentor-src\dsh\install.ps1"; Remove-Item -Recurse -Force "$env:TEMP\blender-mentor-src"

Linux / macOS / WSL:

tmp="$(mktemp -d)"; git clone --depth 1 https://github.com/Memory175/blender-mentor.git "$tmp/bm" && bash "$tmp/bm/dsh/install.sh"; rm -rf "$tmp"

Alternative: download the installer from the latest Release (if release assets are reachable):

irm https://github.com/Memory175/blender-mentor/releases/latest/download/dsh-install.ps1 -OutFile "$env:TEMP\dsh-install.ps1"; powershell -NoProfile -ExecutionPolicy Bypass -File "$env:TEMP\dsh-install.ps1"

From a clone: powershell -File dsh/install.ps1 / bash dsh/install.sh.

DSH (plugin market)

This repository is also a DSH skill plugin: package.json declares dsh.bundle.patch → cordis.patch.yml, which mounts the bundled skills/blender-mentor/ as a filesystem skill provider. After the dsh.so listing is approved, install it with:

dsh plugin --profile web add github:Memory175/blender-mentor

Claude Code

Place it in ~/.claude/skills/blender-mentor/ (user level) or .claude/skills/blender-mentor/ (project level).

Or install it as a plugin from this repository's marketplace:

/plugin marketplace add Memory175/blender-mentor
/plugin install blender-mentor@blender-mentor-marketplace

Claude.ai (Skills-enabled plans)

Zip the repo (or upload SKILL.md together with its directories) and add it to your Skills library.

Cursor

Open this repo in Cursor and reference it in chat with @SKILL.md / @references/main.md. To keep it always on, put SKILL.md under .cursor/rules/.

Verify

python -m compileall -q scripts tools
python scripts/persona_lint.py . --allow-no-deny   # structure gate
python scripts/selftest.py .                       # toolchain self-test (14 pass + 1 skipped)

Blender-side checks (requires Blender 4.2 LTS / 5.x):

blender -b file.blend --python scripts/asset_gate.py -- --json gate.json
blender -b scene.blend --python scripts/biomech_check.py -- --eccentric 1,4 --support 1-4

Usage examples

Use blender-mentor to walk me through M1 modeling and topology
Review this walk cycle: grade it with the rubric and give a P0/P1/P2 fix list
Turn this bounce animation's deformation envelope into an acceptance-testable spec
Why is Blender throwing this error? Start with an L2 hint

Versioning & updates

  • Suite version: V5.4.0; skill package release: v5.4.2 (see CHANGELOG.md);
  • After editing anything in references/, re-run persona_lint.py;
  • Rebuild the skill directory: python tools/build_skill.py --source <V5.4.0 six-file suite> --out <repo parent> --force (keeps .git).

License

MIT — see LICENSE. Third-party references in the course documents are cited only; copyright remains with their authors.