architecture-map MCP server
Enables MCP-compatible AI agents to read and write architecture-map projects and diagrams with per-project access controls via OAuth 2.1/PKCE.
README
Cartogram MCP server
Status: early-stage, active development. Issues and PRs are welcome.
Remote MCP server (Streamable HTTP + OAuth 2.1/PKCE) that lets an
MCP-compatible AI agent (e.g. Claude Code) read and write a user's
Cartogram projects and diagrams, subject to the same per-project
access grants and read/write/admin scopes the user controls from
/settings/integrations in the web app.
Production instance: https://mcp.cartogram.andrewpg.me — point any
MCP-compatible client at that URL, it self-registers via Dynamic Client
Registration (RFC 7591) and walks the OAuth consent screen from there.
Local development
Requires the Supabase backend running — see the sibling
cartogram-supabase repo:
supabase start # from the cartogram-supabase repo
Then, from this repo:
cp .env.example .env
# fill in SUPABASE_URL / SUPABASE_ANON_KEY / SUPABASE_SERVICE_ROLE_KEY
# from `supabase start`'s output, a random MCP_JWT_SIGNING_SECRET, and
# MCP_FRONTEND_URL (where the frontend's OAuth consent screen lives)
npm install
npm run dev
Testing
npm test
Tools exposed
| Tool | Scope | Notes |
|---|---|---|
list_projects |
read | |
list_diagrams |
read | returns diagrams no other diagram's childDiagram reaches — the ones a picker/agent needs a direct slug for |
get_diagram |
read | |
create_project |
write | auto-grants MCP access to the created project; seeds an empty deployment diagram |
create_diagram |
write | |
update_diagram |
write | optimistic-locking: pass the version from a prior get_diagram call; a mismatch returns { conflict: true } |
validate_diagram |
none | dry-run shape validation, no DB write |
invite_collaborator |
admin |
Known limitations
delete_projectandremove_collaboratorare intentionally not exposed (destructive, low-value for the "document a repo" flow this is built around — add later if there's real demand).childDiagramslugs are not validated against existing diagrams at write time — a typo creates a dangling reference instead of failing fast.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.