Skip to content

design:os · 3D Blender · agent operating system

An AI agent that builds parts natively in Blender, and shows its evidence.

Agent skills, a verified bpy knowledge base, an AGENT_OK / AGENT_FAIL execution contract and a production gate for Blender 5.2 LTS. Every worked build below ships with what was checked and what was not.

License: PolyForm Shield 1.0.0 Last commit Blender 5.2 LTS Python 3.10+ bpy: data API first MCP: blender-mcp Claude Code · Codex · Antigravity

DESIGN:OS 3D Blender in 30 seconds: five worked builds going from blueprint lines to clay to the final render

rendered with the DESIGN:OS SVG Animation engine, 1080p 60 fps · Download the 9:16 version

Worked builds

Names, status and evidence limits live next to each build. ORC images are derivatives of native renders whose source models are not in the public checkout; the keyboard, robot arm and watch winder are.

  • Robot arm, overview render

    Robot arm

    Native assembly film, sampled motion checks, exported fit prototypes and a 38-part digital gate report. Loaded operation and physical assembly stay separate.

    Gallery on GitHub
  • Watch-winder capsule, overview render

    Watch-winder capsule

    Native stills, sampled motion checks, digital print-kit checks and fit visualization. Printing and physical fit stay separate.

    Gallery on GitHub
  • Wellhead, overview render

    Wellhead

    Overall and valve details; render review imagery, physical qualification is separate.

    Gallery on GitHub
  • Separator, overview render

    Separator

    Vessel, nozzles and instrumentation; render-only detail set including the repaired vapor interface.

    Gallery on GitHub
  • Exchanger train, overview render

    Exchanger train

    Overview, bellows, instruments and hardware; current R7 HQ imagery with stated technical limits.

    Gallery on GitHub
  • Three-stage exchanger, overview render

    Three-stage exchanger

    Circuit layout and installed instruments; verification previews, HQ delivery in progress.

    Gallery on GitHub
  • Turbine–generator, overview render

    Turbine–generator

    Eight HQ overall and close-up views; refreshed cabinet revision, visual review completed.

    Gallery on GitHub
  • Instrument families, overview render

    Instrument families

    Pressure and temperature mounts shared across assemblies; reuse needs a local origin, a mounting interface and verification in the receiving assembly.

    Gallery on GitHub

Install and use

Requirements: Blender 5.2.x, Python 3.10+, macOS or Linux. The host-side tests use the Python standard library only.

0 · Model Context Protocol (MCP) — For Humans & AI

Instant zero-install setup via uvx. Equips Claude Desktop, Cursor, Antigravity, and Cline with verified Blender 5.2 knowledge, mechanical spec validation, and headless pass execution.

# Zero-install execution (runs anywhere without local setup):
uvx design-os-3d-blender

# Client configuration (Claude Desktop, Cursor, Antigravity, Cline):
{
  "mcpServers": {
    "design-os-3d-blender": {
      "command": "uvx",
      "args": ["design-os-3d-blender"],
      "env": {
        "BLENDER_BIN": "/Applications/Blender.app/Contents/MacOS/Blender"
      }
    }
  }
}

1 · check the install

Point the scripts at your Blender binary, verify the knowledge catalogue, run the host tests.

git clone https://github.com/jangtrinh/design-os-3d-blender.git && cd design-os-3d-blender
export BLENDER_BIN=/Applications/Blender.app/Contents/MacOS/Blender
python3 scripts/blender-knowledge.py check
python3 -m unittest discover -s tests/execution

2 · run a pass headless

Every pass runs in an isolated Blender process and ends with a sentinel; success is AGENT_OK on the last line, never the exit code.

bash scripts/headless-run.sh my-pass.py

3 · or interactively over MCP

Open the blender-mcp addon in Blender’s sidebar (N → BlenderMCP → Connect), then run the same pass through the runtime from your agent CLI.

import sys; sys.path.insert(0, "<ROOT>/scripts")
import agent_runtime as rt
rt.run_file("<ROOT>/builds/<slug>/pass-01.py")

4 · gate a printable part

Write the spec before detailing; the gate checks the scene against it and writes a report. A digital gate does not establish physical fit, load, pressure, thermal duty or a successful print.

mkdir -p builds/my-part && cp specs/examples/bracket-m3.spec.json builds/my-part/spec.json
python3 scripts/production-gate.py --scene builds/my-part/part.blend --spec builds/my-part/spec.json --report builds/my-part/gate-report.json

Tools and skills

Model Context Protocol Server

FastMCP server exposing runtime inspection, verified bpy 5.2 knowledge base search, and spec validation tools with zero Blender dependency.

Core Blender skill

Native scene execution, numerical checks and delivery recipes. blender-agent-core

Image reconstruction skill

Reference contracts, staged modeling and construction-aware visual review. blender-image-to-3d

Knowledge workbench

Bounded reading packs and reviewed source routing over the verified bpy knowledge base.

Native pass pipeline

Declared step dependencies, durable attempt journals and hash-checked resumption; execution status stays separate from acceptance.

Verification and production gate

Scene checks, isolated previews, spec-bound digital geometry and export evidence; HQ coverage receipts before a guarded render.

MakerWorld publishing leg

Bambu project validation and the documented publishing constraints.

DESIGN:OS Ecosystem

Modular toolchain for AI coding agents: from interface CLI to deterministic animation, 3D CAD and robotics.

Read next

Watch

Three videos from the design:os by Jang channel. Thumbnails are self-hosted; the player loads only when you press play.

  • How to use design os 3d blender repo - ep 1 (YouTube, 2:57)2:57

    Ep 1: CK-001 keyboard

  • How to use design os 3d blender repo - ep 2 (YouTube, 5:34)5:34

    Ep 2: DC-01 desk robot

  • Fanmade 3D model for Nowa planet - by design os 3d blender (YouTube, 0:27)0:27

    Fan-made Nowa model (Short)

More on the channel

Listed on

  • Listed on mcpservers.org
  • Featured AI Agents on AI Agents Directory