Generate UML and other diagrams through the Model Context Protocol. Supports 30+ diagram types through Kroki, PlantUML, Mermaid, and D2.
Live endpoint: https://uml-mcp.vercel.app/mcp Smithery: Add via Smithery
- Remote (recommended): Fast setup over HTTP MCP with Vercel serverless runtime
- Local: stdio process for file output and local debugging
Use this when you want the deployed endpoint from this repo:
"uml-mcp": {
"transport": "http",
"url": "https://uml-mcp.vercel.app/mcp"
}git clone https://github.com/antoinebou12/uml-mcp.git && cd uml-mcp
uv sync
uv run python server.pyFor client config snippets, use:
config/cursor_config.jsonconfig/claude_desktop_config.jsonconfig/README.mdfor exact config file locations
- Transport: Remote uses HTTP MCP, local uses stdio by default
- Runtime: Remote runs on Vercel, local runs in your Python environment
- File writes: Remote is read-only (no
output_dir), local supportsoutput_dir - Returned data: Both return URL + base64; local can also save files
- Environment variables: Remote is managed server-side; local reads your env config
Important: The MCP route is /mcp, not the domain root.
- 30+ diagram types -- UML (Class, Sequence, Activity, Use Case, State, Component, Deployment, Object), Mermaid, D2, Graphviz, TikZ, ERD, BlockDiag, BPMN, C4, and more via Kroki
- MCP-native tools --
generate_uml,validate_uml(optionalstrict),list_diagram_types,generate_uml_batch - Multiple outputs -- SVG, PNG, PDF, JPEG, and base64 (varies by diagram type)
- Fallback pipeline -- Kroki first, then PlantUML or Mermaid.ink
- Flexible deployment -- local stdio, local HTTP, Docker, Vercel, Smithery
- UML (PlantUML): Class, Sequence, Activity, Use Case, State, Component, Deployment, Object
- General: Mermaid, D2, Graphviz, ERD, BlockDiag, BPMN, C4
- Specialized: TikZ, Excalidraw, Nomnoml, Pikchr, Structurizr, SVGBob, WaveDrom, WireViz, and more
Full list with supported formats: run python server.py --list-tools or query uml://formats.
generate_uml-- render a diagram; omitoutput_dirfor URL/base64 onlyvalidate_uml-- structural validation before render (strictfor extra Mermaid/D2 checks)list_diagram_types-- same metadata asuml://typeswhen resources are awkwardgenerate_uml_batch-- multiple diagrams in one call (cap:MCP_BATCH_MAX_ITEMS)
types, templates, examples, formats, capabilities, server-info, mermaid-examples, bpmn-guide, workflow
This repo includes vercel.json for serverless deployment.
- Connect the repo to Vercel
- Use
https://<project>.vercel.app/mcp - Keep
/mcpin all MCP client URLs
- Open smithery.ai/new, choose URL
- Enter
https://<project>.vercel.app/mcp - Configure display name, description, and homepage
Detailed guide: docs/integrations/vercel_smithery.md
Default image serves FastAPI on port 8000 with MCP HTTP at http://127.0.0.1:8000/mcp.
# Full local stack (local Kroki + mermaid + blockdiag)
docker compose up -d
# API + MCP only (public Kroki)
docker build -t uml-mcp . && docker run -p 8000:8000 uml-mcp
# stdio MCP subprocess mode
docker run -i uml-mcp python server.py --transport stdioThese variables apply to local/self-hosted runs. Remote Vercel endpoint settings are managed server-side.
KROKI_SERVER-- Kroki server URL (default:https://kroki.io)PLANTUML_SERVER-- PlantUML server URL (default:http://plantuml-server:8080)MCP_OUTPUT_DIR-- diagram output directory (default:./output)MCP_READ_ONLY-- disable file writes (default:false)MCP_MAX_CODE_LENGTH-- max diagram code length (default:500000)MCP_BATCH_MAX_ITEMS-- max items pergenerate_uml_batch(default:20)MCP_RATE_LIMIT_PER_MINUTE-- HTTP rate limit per IP for diagram/MCP routes;0= off (default:0)USE_LOCAL_KROKI-- use local Kroki instance (default:false)USE_LOCAL_PLANTUML-- use local PlantUML instance (default:false)
Full options: docs/configuration.md
server.py -- MCP entry point (stdio/HTTP)
app.py -- FastAPI REST API + MCP HTTP at /mcp
api/app.py -- legacy re-export of root app (Vercel FastAPI preset uses root app.py)
mcp_core/
core/ -- config, server, CLI, utilities, diagram pipeline
tools/ -- generate_uml, validate_uml
prompts/ -- diagram generation prompts
resources/ -- uml:// resource handlers
tools/kroki/ -- Kroki, PlantUML, Mermaid, D2 clients
Fallback strategy: Kroki (primary) -> PlantUML server (UML types) / Mermaid.ink (Mermaid) -> error with details.
# Install dev dependencies
uv sync --all-groups
# Run tests
uv run pytest tests/ -v
# Lint
uv run ruff check . && uv run ruff format --check .
# Integration tests (requires real FastMCP)
USE_REAL_FASTMCP=1 uv run pytest tests/integration -v
# Local CI
make ci- Online: antoinebou12.github.io/uml-mcp
- Local:
uv run mkdocs servethen open http://127.0.0.1:8000