Diagrams MCP Server
Enables creation of infrastructure and architecture diagrams as code using the Python diagrams library, supporting 15+ cloud providers, 500+ node types, custom icons, and flowcharts with multiple output formats.
README
Diagrams MCP Server
MCP server for generating infrastructure and architecture diagrams as code using the Python diagrams library.
Features
5 Diagram Tools for infrastructure, architecture, and flowcharts:
- Infrastructure Diagrams - 15+ providers (AWS, Azure, GCP, K8s, On-Prem, SaaS)
- 500+ Node Types - Compute, database, network, storage, security
- Custom Icons - Web URLs (HTTPS) and local files
- Flowcharts - 24 shapes for process diagrams
- Validation - Dry-run before generation
Advanced Capabilities:
- Multiple output formats (PNG, PDF, JPG, DOT)
- Cluster grouping with unlimited nesting
- Edge styling (colours, labels, line styles)
- Graphviz attribute customisation
Installation
System Requirements:
- Graphviz must be installed:
- macOS:
brew install graphviz - Ubuntu/Debian:
sudo apt-get install graphviz - Windows: Download from https://graphviz.org/download/
- macOS:
IDEs
Claude Desktop
Add to your claude_desktop_config.json:
For published package:
{
"mcpServers": {
"diagrams": {
"command": "uvx",
"args": ["diagrams-mcp"]
}
}
}
For local development:
{
"mcpServers": {
"diagrams:local": {
"command": "uv",
"args": [
"--directory",
"/ABSOLUTE/PATH/TO/diagrams-mcp",
"run",
"diagrams-mcp"
]
}
}
}
Claude Code
Quick setup (CLI):
Published package:
claude mcp add --transport stdio math -- uvx diagrams-mcp
Local development:
claude mcp add --transport stdio math -- uvx --from /absolute/path/to/diagrams-mcp diagrams-mcp
Team setup (project-level):
Add .mcp.json to your project root:
{
"mcpServers": {
"diagrams": {
"command": "uvx",
"args": ["diagrams-mcp"]
}
}
}
Verify installation:
claude mcp list
Or check in IDE: View → MCP Servers, or use /mcp command.
Try It
Once installed, try these prompts:
- "Create an AWS 3-tier web application diagram with Route53, ELB, EC2 instances, and RDS"
- "Generate a Kubernetes microservices architecture with ingress, services, and pods"
- "Build a flowchart for a CI/CD pipeline with decision points"
- "Create a diagram using a custom icon from my company logo URL"
- "Show me all available AWS compute nodes"
Map to tools: create_diagram, create_diagram_with_custom_icons, create_flowchart, list_available_nodes, validate_diagram_spec
Tool Reference
All tool parameters and descriptions are available in your IDE's autocomplete.
Diagram Generation (3 tools)
| Tool | Description |
|---|---|
create_diagram |
Full infrastructure/architecture diagrams with all providers |
create_diagram_with_custom_icons |
Diagrams with custom node icons from URLs or local files |
create_flowchart |
Simplified flowchart creation with 24 process shapes |
Discovery & Validation (2 tools)
| Tool | Description |
|---|---|
list_available_nodes |
Search 500+ available nodes by provider, category, or keyword |
validate_diagram_spec |
Dry-run validation before generation |
Custom Icons
Web URL Icons
- HTTPS-only (HTTP rejected)
- 5MB file size limit
- 5-second download timeout
- Image format validation (PNG, JPG)
- Automatic caching (~/.diagrams_mcp/icon_cache)
Local File Icons
- Path validation (file must exist)
- Format validation
- Sandboxed execution
Development
Running Tests
# Run all tests
uv run poe test
Development Modes
STDIO mode (for Claude Desktop integration):
uv run diagrams-mcp
HTTP mode (for containerised deployments):
uv run diagrams-mcp-http --port 8000
License
MIT License. See LICENSE file for details.
Contributing
Contributions welcome via PRs! Please ensure:
- Tests pass, and new ones are added if applicable
- Code is linted & formatted
- Type hints are included
- Clear, actionable error messages are provided
Support
For issues and questions, please open an issue on GitHub.
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。