SEODesk MCP Server
Exposes SEODesk SEO tools (analyze_website, keyword_research, backlink overview, etc.) to MCP clients like ChatGPT via Streamable HTTP and OAuth 2.1.
README
SEODesk MCP Server
A Remote MCP (Model Context Protocol) server that exposes the SEODesk SEO toolset to ChatGPT Apps and any other MCP-compatible client (Claude, Cursor, Codex, etc.).
- Transport: Streamable HTTP (spec 2025-06-18) at
POST /mcp - Auth: OAuth 2.1 with PKCE + Dynamic Client Registration (RFC 7591)
- SDK:
@modelcontextprotocol/sdk - Runtime: Node.js 20+
The Lovable frontend is untouched. ChatGPT talks to this MCP server; the frontend keeps talking to the existing Supabase Edge Functions. This server calls the same Edge Functions internally to fulfil tool calls.
Tools exposed
| Tool | Purpose |
|---|---|
analyze_website |
Full on-page SEO analysis for a URL |
seo_audit |
Technical site audit (findings by category) |
page_speed_analysis |
Google PageSpeed Insights (mobile + desktop) |
keyword_research |
Keywords, questions, clusters from a seed |
generate_meta_title |
AI-generated SEO <title> variants |
generate_meta_description |
AI-generated meta descriptions |
generate_schema_markup |
JSON-LD schema (Article, Product, FAQPage, …) |
backlink_overview |
Backlink profile summary (requires provider key) |
content_optimizer |
Content brief vs. a target keyword |
Endpoints
| Path | Description |
|---|---|
POST /mcp |
MCP Streamable HTTP endpoint (requires Authorization: Bearer <token>) |
GET /health |
Liveness probe |
GET /ready |
Readiness probe |
GET /.well-known/oauth-authorization-server |
OAuth 2.0 AS metadata (RFC 8414) |
GET /.well-known/oauth-protected-resource |
MCP resource metadata (RFC 9728) |
POST /oauth/register |
Dynamic Client Registration (RFC 7591) |
GET /oauth/authorize |
Authorization endpoint (PKCE S256 required) |
POST /oauth/token |
Token endpoint (auth code + refresh) |
Local development
cd mcp-server
cp .env.example .env # fill in values
npm install
npm run dev
Sanity check:
curl http://localhost:3333/health
curl http://localhost:3333/.well-known/oauth-authorization-server
Environment variables
See .env.example. Required at minimum:
PUBLIC_BASE_URL— public HTTPS URL of this server (used as OAuth issuer).JWT_SECRET— long random string, signs access tokens.SUPABASE_URL,SUPABASE_ANON_KEY,SUPABASE_SERVICE_ROLE_KEY— to call SEODesk backend Edge Functions.LOVABLE_API_KEY— for AI-powered tools (meta/schema/optimizer).GOOGLE_PAGESPEED_API_KEY— optional (Edge Function can supply its own).
Deploy to Railway
- Push this folder to a Git repo.
- In Railway → New Project → Deploy from GitHub, pick the repo and set the
root directory to
mcp-server. - Add the env vars from
.env.exampleunder Variables. - Railway auto-detects Node;
railway.tomlsets the healthcheck to/health. - After deploy, set
PUBLIC_BASE_URLandOAUTH_ISSUERto the generated HTTPS domain (e.g.https://seodesk-mcp.up.railway.app) and redeploy.
Deploy to Cloudflare Workers
The MCP SDK targets Node's http module, so the simplest CF path is to deploy
this Node app to Cloudflare Containers (or Fly.io / Render) rather than a
plain Worker. If you want a pure Workers deployment, port the two files
src/index.js and src/oauth.js onto itty-router + @modelcontextprotocol/sdk's
Web-Standard transport — the tool definitions in src/tools.js transfer
unchanged.
Minimal Docker deploy (works on Fly.io, Render, CF Containers):
docker build -t seodesk-mcp .
docker run -p 3333:3333 --env-file .env seodesk-mcp
Connect from ChatGPT
- ChatGPT → Settings → Apps → Create App (or Connectors).
- Server URL:
https://<your-domain>/mcp - ChatGPT auto-discovers
/.well-known/oauth-protected-resource, registers a client via/oauth/register, and walks the user through OAuth. - Once approved, all nine tools appear in ChatGPT and can be invoked in chat.
Connect from Claude / Cursor / Codex
Any MCP client that supports Streamable HTTP + OAuth 2.1 works. Point it at
https://<your-domain>/mcp and it will complete DCR + PKCE the same way.
Production hardening checklist
- [ ] Replace in-memory
clients/authCodes/refreshTokensstores insrc/oauth.jswith Redis or Postgres. - [ ] Replace the auto-approve
/oauth/authorizehandler with a real consent page that authenticates the end user against SEODesk (Supabase Auth). - [ ] Rotate
JWT_SECRETand pinALLOWED_REDIRECT_HOSTS. - [ ] Put the server behind HTTPS (Railway / Fly / CF terminate TLS for you).
- [ ] Add rate limiting (e.g.
express-rate-limit) on/mcpand/oauth/*.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.