mcp-server-diagramai

Name: mcp-server-diagramai
Author: PadsterH2012

@PadsterH2012

リポジトリ

mcp-server-diagramai について

概要はまだありません

#mcp-server-diagramai #diagram-ai #real-time-diagram-management

基本情報

カテゴリ

その他

ライセンス

MIT license

ランタイム

node

トランスポート

stdio

公開者

PadsterH2012

設定

標準の設定はありません

このサーバーの README には解析可能な MCP 設定ブロックが含まれていません。インストール手順はリポジトリをご確認ください。

リポジトリ

ツール

Create new diagrams with specified content

Retrieve diagram by ID

Modify existing diagrams

Remove diagrams

Add nodes to diagrams

Add connections between nodes

Get all accessible diagrams

Check diagram state and metadata

概要

What is mcp-server-diagramai?

mcp-server-diagramai is an MCP server for DiagramAI integration that lets AI agents create, read, edit, and delete diagrams in real-time. It is designed for developers building AI‑powered diagramming workflows.

How to use mcp-server-diagramai?

Install globally via npm (npm install -g mcp-server-diagramai) or build from source. Configure the server with environment variables for DiagramAI API URL and API key, then add it to the Claude Desktop configuration. Use the exposed tools (e.g., create_diagram, update_diagram) to manage diagrams through your AI agent.

Key features of mcp-server-diagramai

Exposes diagram list and project structure as resources.
Real-time diagram state access and project-based organization.
Tools for full CRUD on diagrams and projects.
Node and edge operations for dynamic diagram editing.
Supports reactflow and mermaid diagram types.
Environment‑variable configuration and debug logging.

Use cases of mcp-server-diagramai

An AI agent automatically creates flowcharts or network diagrams.
An agent modifies existing diagrams in real‑time during a conversation.
An agent reads diagram content to answer user questions.
An agent deletes obsolete diagrams to keep projects clean.

FAQ from mcp-server-diagramai

What environment variables are required?

DIAGRAMAI_API_URL (the URL of your DiagramAI instance) and DIAGRAMAI_API_KEY (with permissions like diagrams:create, diagrams:read, diagrams:update, diagrams:delete, and projects:read). Optionally set REQUEST_TIMEOUT, DIAGRAMAI_WS_URL, and DEBUG.

How do I install mcp-server-diagramai?

Install via npm: npm install -g mcp-server-diagramai. Alternatively, clone the repository, run npm install && npm run build, then point to the built dist/index.js.

How do I configure it with Claude Desktop?

Add a diagramai entry to your Claude Desktop config (mcpServers). For the published package, use command: "npx" with args ["mcp-server-diagramai", "--api-url", "${DIAGRAMAI_API_URL}", "--api-key", "${DIAGRAMAI_API_KEY}"]. For development, provide the full path to Node and the built script plus env variables.

What tools are available?

The server provides create_diagram, get_diagram, update_diagram, delete_diagram, add_node, add_edge, list_diagrams, and get_diagram_status. Each accepts parameters documented in the README.

I get an authentication error. What should I check?

Verify that the API key starts with da_, has the required permissions in DiagramAI settings, and has not expired. Also confirm the API URL and port are correct and the DiagramAI instance is running.