Print-OS

Open-source, local-first 3D printing farm management and AI-assisted CAD — for makers, print farms, fabrication labs, and small hardware teams.

What it is

Print-OS is a 3D printing operating system for studios that run more than one printer. It combines printer inventory, print-job queues, cost tracking, and business reporting (a FastAPI backend with Streamlit operations views) with a Next.js + Three.js workspace for AI-assisted parametric CAD. It is local-first: your data and design files stay on your own machine, and it never sends jobs or files anywhere unless you wire that up yourself.

Architecture

Print-OS is a monorepo of three cooperating surfaces:

• caedo-api — FastAPI backend for business logic, smart job routing, and cost calculation.
• caedo-web — Next.js 16 + React frontend for the CAD workspace and analytics.
• modules/3d-designer — standalone AI-assisted parametric 3D design workspace (Three.js / JSCAD).

Install / Quick start

# 1. Configure environment
cp env.example .env
# edit .env with your API keys

# 2. Start the backend (FastAPI)
cd caedo-api
PYTHONPATH=. python3 api/main.py

# 3. In a second terminal, start the frontend (Next.js)
cd caedo-web
npm install
PORT=3002 npm run dev   # http://localhost:3002 (the port the smoke test checks)

Usage

Verify a working install with the unified smoke test from the repo root:

bash scripts/smoke-test.sh

For exact backend, Streamlit, web, 3D-designer, security, and release checks, see the Operator Runbook. Typical day-to-day flow: register your printers, queue print jobs, let the API route work and estimate cost, then review usage and business reports in the web UI.

Agent surfaces

Print-OS now ships direct agent/operator entrypoints:

• CLI: python3 tools/print_os_cli.py manifest --format json
• MCP: .mcp.json registers tools/print_os_mcp.py as a local stdio MCP server.
• Skill: skills/print-os/SKILL.md gives Codex, Claude Code, and other skill-aware agents the safe operating workflow.

Useful CLI examples:

python3 tools/print_os_cli.py job-payload --name "Bracket" --material PLA --width 40 --depth 20 --height 12
python3 tools/print_os_cli.py api-get health --base-url http://127.0.0.1:8000
python3 tools/print_os_cli.py create-job --name "Bracket" --material PLA --width 40 --depth 20 --height 12

The MCP server exposes print_os_manifest, print_os_job_payload, print_os_api_get, and print_os_create_job. Read operations are safe; print_os_create_job writes to the local running API and should be treated as an explicit operator action.

Why / how it works

Most maker tooling stops at slicing a single part. Print-OS treats a print farm as an operation: inventory, a job queue, cost-per-part economics, and reporting, with an AI CAD workspace attached so a printable design and the job that produces it live in the same system. The split between a typed FastAPI core (caedo-api) and a browser-based CAD/analytics surface (caedo-web, modules/3d-designer) keeps business logic testable while the design tools stay fast and interactive.

Stack: FastAPI, Streamlit, Next.js 16, React, Three.js, JSCAD, SQLite, Python, TypeScript.

Best-fit searches

3D printing farm management software · open-source print farm manager · maker studio production planning · print job queue tracker · printer inventory and cost tracking · AI-assisted CAD for 3D printing · local-first fabrication operations · small-batch hardware prototyping

Links

• AI / agent navigation: llms.txt
• Agent skill: skills/print-os/SKILL.md
• Local MCP config: .mcp.json
• Operator verification: docs/OPERATOR_RUNBOOK.md
• License: MIT
• KyaniteLabs: kyanitelabs.tech
• Sibling projects: GameStory-Lab · voice-to-sculpture-app · grocery-flywheel · HealthAdvocate · CyberWitches