dcc-mcp-marvelous-designer
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@dcc-mcp-marvelous-designerImport the ZPRJ and simulate it for 30 steps"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
dcc-mcp-marvelous-designer
Typed garment operations for Marvelous Designer, using DCC-MCP Core discovery, execution and lifecycle. Initial integration targets the documented 2025 Python API; no licensed host or engine acceptance has been completed.
Supported API contracts
Read API availability and paginated pattern information.
Import ZPRJ without the general import dialog; this replaces the current project.
Create a rectangular pattern and verify the pattern count.
Simulate 1–30 synchronous steps. The documented API has no verified interrupt operation. A timeout is not proof simulation stopped; do not retry blindly.
Save to a new ZPRJ; export OBJ, FBX or offline Alembic into an empty directory.
Hash and check the actual returned nonempty artifacts, restricted to configured roots.
The adapter adds no arbitrary Python, shell, HTTP proxy, login, activation or UI input. API presence does not prove a license is available. No vendor files are bundled.
Related MCP server: Daz Pilot
Host bootstrap
Install the package and a compatible official Core wheel in the host Python environment only after operator authorization. Register a Python plugin in the Marvelous Designer Plugin Manager, or use its Python editor. On the host main thread:
from dcc_mcp_marvelous_designer.qt_bootstrap import start
server = start(roots=["C:/garment-project"], dcc_version="YOUR_ACTUAL_VERSION")This optional bootstrap only reuses an already loaded PySide2/PySide6 binding
with a live Qt application. Its availability in Marvelous Designer is unverified.
It refuses unsupported hosts instead of loading another Qt ABI or calling host
APIs on workers. If absent, server.start_server(roots, dispatcher, ...) requires
a verified host-thread dispatcher. Do not use an inline dispatcher for GUI calls.
Stop with qt_bootstrap.stop() before unloading. Core owns registration and the
OS-assigned endpoint. Discover with dcc-mcp-cli list, then search/load the
marvelous-project and marvelous-garment skills for that exact instance.
Validation
python -m pytest
python -m buildTests use deterministic host doubles and prove adapter contracts, not cloth physics, host integration, Unreal or Unity compatibility. See acceptance for the required live chain and gaps.
Sources
Official API, option types, simulation examples, plugin registration.
Independent integration; not affiliated with or endorsed by the software vendor.
This server cannot be deployed
Maintenance
Related MCP Connectors
Control Unreal Engine to browse assets, import content, and manage levels and sequences. Automate…
Generate PDF, Word (.docx) and PowerPoint (.pptx) documents from Markdown over MCP.
Market Data App MCP — wraps the Market Data App API (marketdata.app)
Governed data discovery, exact queries, decisions, simulations, and runtime utilities over MCP.
Related MCP Servers
- AlicenseBqualityDmaintenanceEnables AI-assisted garment design by letting Claude (or any MCP host) drive CLO3D — import projects, dress avatars, assign fabrics, simulate cloth, render, and export, all from a chat.143MIT
- AlicenseNot gradedqualityCmaintenanceMCP server for AI-driven DAZ Studio scene editing. Runs as a native DAZ Studio plugin (DLL) and exposes all scene control, rendering, dForce, asset, and animation commands through the Model Context Protocol.7MIT
- FlicenseNot gradedqualityAmaintenanceEnables MCP-capable agents to control REAPER through safe, typed operations for game audio workflows such as creating variations and rendering WAV files.1-
- AlicenseNot gradedqualityAmaintenanceTyped DCC-MCP control for Marmoset Toolbag 4.03+ and 5.x, enabling shell-capable agents to interact with Toolbag scenes via a shared CLI and MCP gateway.2MIT