modelscope·@Noacodenoobe/drawioMCP
暂无描述。
A Model Context Protocol (MCP) server that provides programmatic tools for creating and managing draw.io diagrams using mxgraph. Generate architecture diagrams, flowcharts, and other visualizations through a clean API that works with Claude Desktop and other MCP-compatible clients.
This server enables you to build diagrams incrementally by providing stateless tools that operate on .drawio.svg files. Each operation specifies the target file, making it compatible with VSCode's draw.io extension while maintaining a clean separation between diagram state and server operations.
.drawio.svg files that work seamlessly with VSCode draw.io extension
Add this configuration to your MCP client (e.g., Claude Desktop, Cursor):
{
"mcpServers": {
"drawio-diagrams": {
"command": "npx",
"args": ["drawio-mcp"]
}
}
}
The server supports both absolute and relative file paths:
/Users/username/project/diagrams/architecture.drawio.svg./diagrams/architecture.drawio.svg (when cwd is configured)All diagram files should use the .drawio.svg extension for proper VSCode integration.
All primary tools support batch operations, allowing you to perform multiple actions in a single MCP call for improved efficiency:
add_nodes: Create multiple nodes simultaneouslyedit_nodes: Update multiple nodes/edges simultaneouslylink_nodes: Create multiple connections simultaneouslyremove_nodes: Remove multiple nodes simultaneouslyThis approach reduces network overhead and provides atomic operations - either all changes succeed or none are applied.
Create a new empty diagram file.
Parameters:
file_path (string, required): Path for the new diagram fileExample:
{
"file_path": "./diagrams/system-architecture.drawio.svg"
}
Add one or more nodes to an existing diagram in a single operation. Optionally run an automatic layout after insertion.
Parameters:
file_path (string, required): Path to the diagram filelayout (object, optional): Automatic layout configuration
algorithm (string, required if layout is provided): One of hierarchical, circle, organic, compact-tree, radial-tree, partition, stackoptions (object, optional): Algorithm-specific options
hierarchical only: direction "top-down" | "left-right" (default: "top-down")nodes (array, required): Array of node objects to add, each containing:
id (string, required): Unique identifier for the nodetitle (string, required): Display label (supports newlines with \n)x (number, required): X coordinate positiony (number, required): Y coordinate positionkind (string, required): Node shape typeparent (string, optional): Parent node ID (default: "root")width (number, optional): Custom widthheight (number, optional): Custom heightcorner_radius (integer, optional): Corner radius in pixels ( 1). Only applies to RoundedRectangle. Default is 12 when kind is RoundedRectangle and corner_radius is omitted. The effective visual radius is capped by draw.io/mxGraph to at most half of the shorter side of the node.Available Node Types:
Rectangle: Standard rectangular nodeEllipse: Oval-shaped nodeCylinder: Database/storage representationCloud: Cloud service representationSquare: Square with fixed aspect ratioCircle: Circular nodeStep: Process step shapeActor: UML actor (stick figure)Text: Text-only nodeRoundedRectangle: Rectangle with rounded corners (supports corner_radius in pixels)Example (Single Node):
{
"file_path": "./diagrams/system-architecture.drawio.svg",
"nodes": [
{
"id": "user-service",
"title": "User Service\nAPI Layer",
"kind": "Rectangle",
"x": 100,
"y": 150,
"width": 120,
"height": 80
}
]
}
Example (Multiple Nodes):
{
"file_path": "./diagrams/system-architecture.drawio.svg",
"nodes": [
{
"id": "user-service",
"title": "User Service",
"kind": "Rectangle",
"x": 100,
"y": 150
},
{
"id": "database",
"title": "Primary DB",
"kind": "Cylinder",
"x": 300,
"y": 150
},
{
"id": "cache",
"title": "Redis Cache",
"kind": "Cylinder",
"x": 200,
"y": 300
}
]
}
Example (With Layout):
{
"file_path": "./diagrams/system-architecture.drawio.svg",
"layout": {
"algorithm": "hierarchical",
"options": { "direction": "left-right" }
},
"nodes": [
{ "id": "api", "title": "API", "kind": "Rectangle", "x": 40, "y": 40 },
{ "id": "service", "title": "Service", "kind": "Rectangle", "x": 200, "y": 40 },
{ "id": "db", "title": "DB", "kind": "Cylinder", "x": 360, "y": 40 }
]
}
Note: The layout runs once after all insertions and considers existing edges in the diagram file. For best results when edges are created or modified lat…
一个为大型语言模型(LLM)设计的、高可用的12306余票查询工具服务,现已搭载智能会话管理引擎。 12306-MCP-Server 将复杂的12306余票查询接口封装为符合 Model Context Protocol (MCP) 规范的工具集,允许AI Agent通过自然语言无缝查询实时火车票、中转换乘和经停站信息。 从 v1.0.0 版本开始,项目引入了全新的智能会话管理系统,通过会话池、动态User-Agent轮换和自动错误恢复机制,将服务的稳定性与反屏蔽能力提升至全新高度。
阿里云可观测 MCP服务,提供了一系列访问阿里云可观测各产品的工具能力,覆盖产品包含阿里云日志服务SLS、阿里云应用实时监控服务ARMS、阿里云云监控等,任意支持 MCP 协议的智能体助手都可快速接入。
自动从文本中识别并提取法律条文,例如在合同、案件描述或法律文书中,精准定位涉及的具体法律条文(如“《民法典》第585条”),并将文本中的法律问题与对应法条内容关联,提取法规名称的服务。
用于哔哩哔哩API的MCP服务器,支持视频搜索、用户内容获取等多种操作,适用于哔哩哔哩内容管理和数据分析场景。
看起来您想要翻译一份英文技术文档,但您提供的内容中只有"asd"这个字符串,并没有实际的文档内容。请提供需要翻译的具体英文技术文档文本,这样我才能帮助您进行准确的翻译。如果有特定部分或格式要求,请一并告知。
一座通过串行通信将物理硬件设备与人工智能大语言模型连接起来的桥梁,允许用户使用自然语言命令控制硬件。