Dcc Mcp Comfyui
FreeNot checkedComfyUI adapter for the DCC Model Context Protocol (MCP) ecosystem — bridges MCP tools to ComfyUI via REST + WebSocket API
About
ComfyUI adapter for the DCC Model Context Protocol (MCP) ecosystem — bridges MCP tools to ComfyUI via REST + WebSocket API
README
Production ComfyUI adapter for the DCC Model Context Protocol (MCP) ecosystem. It gives agents a typed, bounded path from API-format workflow validation to queue execution and artifact delivery through ComfyUI's local REST API.

Real local capture: Blender sphere -> half-mesh revision -> content-addressed publish -> click-to-latest ComfyUI Load3D canvas preview. The reproducible single-node workflow is in docs/showcase/comfyui-load3d-preview.json.
Capabilities
- Workflow validation checks API-format structure, graph references, live node classes, and required inputs before queue submission.
- Catalog discovery exposes bounded features, models, embeddings, node summaries, exact node contracts, and redacted device/runtime status.
- Queue operations inspect IDs without workflow bodies, target one exact prompt for cancellation or history deletion, and verify the resulting state; running cancellation fails closed instead of falling back to ComfyUI's legacy global interrupt.
- Asset handoff uploads bounded images with SHA-256 provenance and atomically downloads only artifacts proven to belong to the requested prompt.
- Artifact discovery is shape-based rather than tied to English labels or an image-only key, so file-shaped image, animation, video, audio, 3D, and custom-node outputs share one bounded contract.
stage_3d_assetpublishes a content-addressed revision from an operator-owned export root and stages it under ComfyUIinput/3dforLoad3D.- The bundled ComfyUI extension adds a canvas-level Update to latest DCC revision action backed by an atomic latest-revision pointer.
- Standalone discovery, packaged Skill subprocesses, and six-part DCC-MCP readiness are supported out of the box.
The production path was live-validated on ComfyUI 0.32.0 with a three-node EmptyImage -> ImageInvert -> SaveImage workflow, including typed validation, execution, status polling, and artifact retrieval.
Quick Start
# Install
pip install dcc-mcp-comfyui
# Start ComfyUI locally (in another terminal)
python main.py --listen 127.0.0.1
# Start the MCP server
dcc-mcp-comfyui --comfyui-base-url http://127.0.0.1:8188
To enable bounded 3D synchronization, configure both trusted roots before starting the adapter:
set DCC_MCP_COMFYUI_SYNC_SOURCE_ROOT=G:\dcc-sync\exports
set DCC_MCP_COMFYUI_INPUT_DIR=G:\apps\ComfyUI\input
dcc-mcp-comfyui --comfyui-base-url http://127.0.0.1:8188
Environment Variables
| Variable | Default | Description |
|---|---|---|
DCC_MCP_COMFYUI_BASE_URL |
http://127.0.0.1:8188 |
ComfyUI server URL |
DCC_MCP_COMFYUI_TIMEOUT |
120 |
Request timeout (seconds) |
DCC_MCP_COMFYUI_SYNC_SOURCE_ROOT |
unset | Trusted producer-export root required by stage_3d_asset |
DCC_MCP_COMFYUI_INPUT_DIR |
unset | Trusted ComfyUI input root required by stage_3d_asset |
DCC_MCP_COMFYUI_SYNC_MAX_ASSET_BYTES |
268435456 |
Maximum accepted 3D asset size |
DCC_MCP_COMFYUI_PORT |
OS-assigned | MCP instance port |
DCC_MCP_GATEWAY_PORT |
9765 |
Gateway port |
DCC_MCP_COMFYUI_ENABLE_GATEWAY_FAILOVER |
true |
Gateway failover |
Skills
The wheel includes four validated Skills and 18 typed tools:
| Skill | Tools | Boundary |
|---|---|---|
comfyui-workflow |
5 | Validate, submit, monitor, resolve prompt-owned artifacts, and stage versioned 3D assets |
comfyui-catalog |
7 | Features, model folders, models, embeddings, node summaries/contracts, and redacted runtime status |
comfyui-queue |
4 | Redacted queue inspection, exact cancellation/history deletion, and memory reclamation |
comfyui-assets |
2 | Bounded image upload and atomic prompt-owned artifact download |
The public surface does not expose arbitrary Python, raw process arguments, local ComfyUI install paths, bulk queue deletion, or unbounded catalog dumps.
MCP Client Config
{
"mcpServers": {
"dcc-mcp-comfyui": {
"url": "http://127.0.0.1:9765/mcp"
}
}
}
Build & Test
uv pip install -e ".[dev]"
uv run pytest
uv run ruff check src/ tests/
Architecture
MCP Client → Gateway (:9765) → ComfyUiMcpServer (OS port)
├─ Workflow scripts → ComfyUIBridge → ComfyUI REST API (:8188)
└─ Asset sync → content-addressed revision → configured ComfyUI input/3d
License
MIT
Installing Dcc Mcp Comfyui
This server has no published package — it is built from source. Open the repository and follow its README.
▸ github.com/dcc-mcp/dcc-mcp-comfyuiFAQ
Is Dcc Mcp Comfyui MCP free?
Yes, Dcc Mcp Comfyui MCP is free — one-click install via Unyly at no cost.
Does Dcc Mcp Comfyui need an API key?
No, Dcc Mcp Comfyui runs without API keys or environment variables.
Is Dcc Mcp Comfyui hosted or self-hosted?
Self-hosted: the server runs locally on your machine via the install command above.
How do I install Dcc Mcp Comfyui in Claude Desktop, Claude Code or Cursor?
Open Dcc Mcp Comfyui on unyly.org, pick your client tab (Claude Desktop, Claude Code, Cursor) and press Install — the config is generated automatically, no JSON editing.
Related MCPs
Fetch
Web content fetching and conversion for efficient LLM usage.
AWS KB Retrieval
Retrieval from AWS Knowledge Base using Bedrock Agent Runtime.
by modelcontextprotocolSpring AI MCP Server
Provides auto-configuration for setting up an MCP server in Spring Boot applications.
llm-analysis-assistant
A very streamlined mcp client that supports calling and monitoring stdio/sse/streamableHttp, and can also view request responses through the /logs page. It also
by xuzexin-hzMCP-Agent
A simple, composable framework to build agents using Model Context Protocol by [LastMile AI](https://www.lastmileai.dev)
by lastmile-aiSpring AI MCP Client
Provides auto-configuration for MCP client functionality in Spring Boot applications.
mcp.natoma.ai
A Hosted MCP Platform to discover, install, manage and deploy MCP servers by [Natoma Labs](https://www.natoma.ai)
MCPHub
Website to list high quality MCP servers and reviews by real users. Also provide online chatbot for popular LLM models with MCP server support.
MCP Servers Rating and User Reviews
Website to rate MCP servers, write authentic user reviews, and [search engine for agent & mcp](http://www.deepnlp.org/search/agent)
mkinf
An Open Source registry of hosted MCP Servers to accelerate AI agent workflows.
Compare Dcc Mcp Comfyui with
Not sure what to pick?
Find your stack in 60 seconds
Author?
Embed badge for your README
Browse similar
All ai MCPs
