Diagrams.so
About
Generate and edit cloud architecture diagrams as native draw.io files, from a prompt or from Terraform
Explore
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Diagrams.soCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"diagrams.so": {
"server": {
"command": "npx",
"args": [
"-y",
"@diagrams-so/mcp"
],
"env": {
"DIAGRAMS_API_KEY": "",
"DIAGRAMS_API_BASE": "",
"DIAGRAMS_API_TIMEOUT_MS": ""
}
}
}
}
}McpServers
{
"server": {
"command": "npx",
"args": [
"-y",
"@diagrams-so/mcp"
],
"env": {
"DIAGRAMS_API_KEY": "",
"DIAGRAMS_API_BASE": "",
"DIAGRAMS_API_TIMEOUT_MS": ""
}
}
}Transport
"stdio"Package
"@diagrams-so/mcp"Registry
"npm"Claude Desktop / Cursor (claude_desktop_config.json/mcp.json)
{ "mcpServers": { "diagrams-so": { "command": "npx", "args": ](https://diagrams.so/create)["-y", "@diagrams-so/mcp"] // optional: "env": { "DIAGRAMS_API_BASE": "http://localhost:8000/api/v2" } for a local API } } }
Connect either by runningnpx @diagrams-so/mcp loginonce, or, withDIAGRAMS_LOGIN_EMAILset, by asking for a diagram and following the link the first tool call gives you (the sign-in code arrives by email). SetDIAGRAMS_API_KEYinstead only for CI and headless machines, where no browser can open.
Restart the client, then ask:"Generate an AWS 3-tier web app diagram and show me the warnings."
Downloaddiagrams-so.mcpbfrom thelatest releaseand drag it into Claude Desktop. No terminal, and it no longer asks for an API key: install it, then connect on first use by clicking the link the first tool call gives you.
Building the bundle yourself is a contributor step, seeDeveloping locally.
The server is listed atsmithery.ai/servers/diagrams-so/mcp, which installs it for you and lists all 23 tools with their parameters:
Same package, sameloginstep. It's the MCPB bundle Smithery installs, not the npm package, so the version shown there follows releases rather thannpm dist-tags.
DIAGRAMS_API_KEY=dgz_live_your_key node test-smoke.mjs
Runs the full flow (connect → list tools → whoami → generate → warnings → fix → export → error handling).
# bump "version" in package.json + manifest.json to match, commit, then: git tag v1.2.0 && git push origin v1.2.0
The tag must matchpackage.json'sversion(the workflow enforces this). The.mcpbbundle appears on the GitHub Release; manualworkflow_dispatchproduces it as a downloadable artifact without publishing (handy for testing a bundle).
Repo secrets/variables (optional):DIAGRAMS_API_KEY(live smoke),NPM_TOKEN(npm publish),DIAGRAMS_API_BASEvariable (non-prod live-smoke target).
- stdout is the MCP channel— the server logs only to stderr.
- Errors come back as clean MCP tool errors carrying the API'scode, HTTP status, andrequest_id.
- The server never talks to internal services or the database — only the public/api/v2.
- Long-running tools stay alive past client timeouts.generate/edit/fix/relayout/enhance_prompt/clarify_promptemit anotifications/progressevery 10s while running, so MCP clients that reset their request timeout on progress (resetTimeoutOnProgress) won't abort a slow generation at the SDK's 60s default. If your client doesn't reset on progress, raise its per-call timeout for these tools.
Only needed if you are changing the server itself. Users should install from npm, seeQuick start.
git clone https://github.com/RedHold/diagrams-mcp-app-core.git cd diagrams-mcp-app-core npm install # installs deps and builds dist/ via the prepare hook node scripts/ci-smoke.mjs # all 23 tools register; no API calls, no credits # point a client at your working copy claude mcp add diagrams-so-dev -- node "$(pwd)/dist/index.js" # build the MCPB bundle npm run build && npx @anthropic-ai/mcpb pack
- Code:Apache-2.0. SeeNOTICE.
- Service:this server is a client for the Diagrams.so API. Use of the API is governed by theTerms of ServiceandAcceptable Use Policy; the code license grants no rights to the API itself.
- Privacy:the server runs locally, connects only toapi.diagrams.so, and contains no telemetry. Credentials fromloginare stored at~/.diagrams-so/credentials.jsonwith owner-only permissions. See thePrivacy Policy.
- Billing:generate, edit, fix, re-layout, and fork operations cost credits; reads and exports are free. Test-mode keys bill your real credit balance.
- Trademarks:Diagrams.so and the Diagrams.so logo are trademarks of RedHold LLC. This license does not grant permission to use them, except to accurately describe the package's origin. See theTrademark Policy.
- Security:report vulnerabilities to[email protected]SECURITY.md.
This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.
Create crafted UI components inspired by the best 21st.dev design engineers.
Bring agent evaluations, observability, and synthetic test set generation directly into your IDE for free with Galileo's new MCP server
An MCP server to help AI assistants to answer questions and generate AccelByte Extend SDK code more effectively .
MCP server for AI Diagram Maker — generate beautiful software engineering diagrams directly inside Cursor, Claude Desktop, Claude Code, or any MCP-compatible AI agent
ALAPI MCP Tools,Call hundreds of API interfaces via MCP
AI-powered SVG animation generator that transforms static files into animated SVG components using the Allyson platform
MCP server that gives AI assistants on-demand access to 1,500+ amCharts docs, ~300 code examples, and 1000+ class API references.
APIMatic MCP Server is used to validate OpenAPI specifications using APIMatic. The server processes OpenAPI files and returns validation summaries by leveraging APIMatic’s API.
One shared context layer for AI agents and humans — live API specs, DB schemas, and versioned contracts across repos so every agent and teammate works from the same source of truth.
Build and deploy full-stack Next.js apps with 98 tools for React, AWS, and MongoDB
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
