PlantUML MCP Server
Enables AI assistants to generate UML diagrams from text descriptions using PlantUML, supporting multiple output formats (PNG, SVG, PDF), syntax validation, and diagram source extraction.
README
PlantUML MCP Server
A Model Context Protocol (MCP) server that wraps PlantUML functionality, allowing AI assistants to generate UML diagrams from text descriptions.
Features
- Generate Diagrams: Create diagrams in multiple formats (PNG, SVG, TXT, PDF, etc.)
- Syntax Checking: Validate PlantUML syntax without generating images
- Source Extraction: Extract PlantUML source from PNG/SVG metadata
- Dark Mode Support: Generate diagrams with dark theme
- Version Info: Get PlantUML and Java version information
Prerequisites
- Node.js 18 or higher
- Java Runtime Environment (JRE)
- PlantUML JAR file
Installation
- Clone this repository:
git clone <repository-url>
cd planuml
- Install dependencies:
npm install
- Build the project:
npm run build
- Set the PlantUML JAR path (optional):
export PLANTUML_JAR=/path/to/plantuml.jar
Default path: /Users/userxxx/bin/plantuml.jar
Usage
As an MCP Server
Add to your MCP client configuration (e.g., Claude Desktop):
{
"mcpServers": {
"plantuml": {
"command": "node",
"args": ["xxx/dist/index.js"],
"env": {
"PLANTUML_JAR": "xxx/plantuml.jar"
}
}
}
}
Available Tools
1. generate_diagram
Generate a diagram from PlantUML source code.
Parameters:
source(required): PlantUML source codeformat(optional): Output format -png,svg,txt,utxt,eps,latex,pdf(default:png)darkMode(optional): Render in dark mode (default:false)
Returns: Base64-encoded diagram data
Example:
{
"source": "@startuml\nAlice -> Bob: Hello\n@enduml",
"format": "svg",
"darkMode": false
}
2. check_syntax
Check PlantUML diagram syntax without generating images.
Parameters:
source(required): PlantUML source code to validate
Returns: Validation result or syntax errors
Example:
{
"source": "@startuml\nAlice -> Bob: Hello\n@enduml"
}
3. extract_source
Extract embedded PlantUML source from PNG or SVG metadata.
Parameters:
filePath(required): Path to the PNG or SVG file
Returns: Extracted PlantUML source code
Example:
{
"filePath": "/path/to/diagram.png"
}
4. get_plantuml_version
Get PlantUML and Java version information.
Parameters: None
Returns: Version information
PlantUML Syntax Examples
Sequence Diagram
@startuml
Alice -> Bob: Authentication Request
Bob --> Alice: Authentication Response
@enduml
Class Diagram
@startuml
class User {
+String name
+String email
+login()
+logout()
}
class Admin {
+manageUsers()
}
User <|-- Admin
@enduml
Activity Diagram
@startuml
start
:Initialize;
if (Check condition?) then (yes)
:Process A;
else (no)
:Process B;
endif
stop
@enduml
Use Case Diagram
@startuml
left to right direction
actor User
actor Admin
rectangle System {
User -- (Login)
User -- (View Profile)
Admin -- (Manage Users)
Admin -- (View Reports)
}
@enduml
Development
Build
npm run build
Watch mode
npm run watch
Environment Variables
PLANTUML_JAR: Path to the PlantUML JAR file (default:/Users/userxxx/bin/plantuml.jar)
Supported Output Formats
- png: PNG images (default)
- svg: SVG vector graphics
- txt: ASCII art diagrams
- utxt: ASCII art with Unicode characters
- eps: Encapsulated PostScript
- latex: LaTeX/TikZ output
- pdf: PDF documents
Error Handling
The server handles various error scenarios:
- Invalid PlantUML syntax
- Missing PlantUML JAR file
- Java runtime errors
- File I/O errors
Errors are returned with descriptive messages to help diagnose issues.
License
MIT
Resources
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。