Bundle
dsh-blender
DeepSeek Harness Blender production, reconstruction, validation, and export plugin
- Source
- CheshireJCat
- stars
- 26 stars
- License
- MIT
- Updated
- Updated yesterday
Readme
# dsh-blender [简体中文](README.zh-CN.md) [](https://github.com/awesome-dsh-plugin/awesome-dsh-plugin) An installable [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness) plugin that gives dsh a complete Blender-backed 3D production workflow. It registers one orchestrator plus 29 directly loadable domain skills, all 26 upstream analysis/validation helpers, and 13 workspace-scoped Blender tools. ## Example output  Camera-free preview generated during a real `dsh-blender` installation smoke test. ## Tools - `blender_status` checks Blender, analysis Python, skills, helpers, and the active workspace. - `blender_scene_info` and `blender_object_info` inspect scenes and object-level geometry, materials, UVs, constraints, shape keys, and animation. - `blender_import` converts supported portable assets into versioned `.blend` sources. - `blender_python` runs a small reviewed `bpy`/`bmesh` chunk and saves a versioned `.blend`. - `blender_preview`, `blender_render`, and `blender_render_frames` create blockout, look-dev, and animation evidence for `read_image` QA. - `blender_export` exports GLB/glTF, FBX, OBJ, STL, USD, or PLY. - `blender_validate_scene` and `blender_validate_export` perform target-aware audits and clean-process re-import checks. - `blender_helper_catalog` and `blender_helper_run` expose all 26 deterministic reference, wireframe, contour, multiview, UV, texture, look, repair, and animation-QA helpers. - The bundled 30-skill stack coordinates blockout, refinement, materials, lighting, cameras, animation, reconstruction, validation, and artifact handoff. No Blender add-on or control port is required: each operation launches a local Blender background process. ## Requirements - DeepSeek Harness `0.1.0-rc.6` - Node.js 20+ - Blender 4.3+ available as `blender` on `PATH`, or configured with an absolute path - Python 3.10+ for the optional reference-analysis helpers ## Install from npm ```bash npx @deepseek-ai/dsh plugin --profile web add dsh-blender npx @deepseek-ai/dsh plugin --profile web exec dsh-blender-setup npx @deepseek-ai/dsh --profile web --dump-config npx @deepseek-ai/dsh web ``` The setup command creates a private `.venv` inside the installed package and installs OpenCV, NumPy, Pillow, and SciPy. Skip it only if you do not need reference, wireframe, multiview, texture, or animation-analysis helpers. To install a pinned GitHub release instead: ```bash npx @deepseek-ai/dsh plugin --profile web add github:CheshireJCat/blender#v0.2.1 ``` ## Develop from source ```bash pnpm install pnpm setup:analysis npx @deepseek-ai/dsh plugin --profile web add . ``` Open [http://127.0.0.1:3080](http://127.0.0.1:3080), create a session whose workspace is the modeling directory, and ask dsh to use `create-3d-model`. See [README.zh-CN.md](README.zh-CN.md) for configuration, security boundaries, headless use, and validation commands.
Install
dsh plugin --profile web add github:CheshireJCat/blender#2e240ed7d939d6b7035075d509fd1d9fb6d57526
Profile: web
With the hub plugin installed, ask your agent to install it by name — it resolves the same plan shown here.
dsh plugin --profile web add github:stvlynn/dsh.fish#path:packages/dsh-plugin-hub
install dsh-blender from the hub