SJU ACM MCP Server

Next.js MCP Server for SJU ACM

This MCP (Model Context Protocol) server provides AI-friendly access to SJU ACM's CMS data through a set of tools and resources. Built with Next.js and the @vercel/mcp-adapter, it enables AI assistants to query, search, and retrieve content from a Contentful space.

Add to Cursor

Go to Cursor Settings -> MCP -> Add MCP Server -> paste the JSON below:

{
  "mcpServers": {
    "SJU-ACM-MCP": {
      "url": "https://sju-mcp.vercel.app/sse"
    }
  }
}

Features

Content Types Supported

• Blog Posts - Articles with rich content, authors, and publish dates
• Meetings - Meeting records with dates, locations, slides, and recordings
• Executive Board Members - Team member profiles with positions and contact info
• Hackathons - Event information with dates, status, and registration links
• Landing Page Graphics - Visual assets for website displays
• Parallax Banners - Interactive banner images with links

Available Resources

contentful-schema

• URI: contentful://schema/content-types
• Description: Complete Contentful content type schemas and field definitions
• Use Case: Understanding the data structure and available fields for each content type

contentful-stats

• URI: contentful://stats/overview
• Description: Real-time statistics and counts for all content types
• Use Case: Getting overview metrics of your Contentful space content

Available Tools

query-blog-posts

Query blog posts with optional filtering

• Parameters:
  • slug (optional): Get a specific post by slug
  • limit (optional): Limit number of results
• Returns: Blog post data with titles, excerpts, authors, and cover images

query-meetings

Query meeting records

• Parameters:
  • type (optional): "all" or "upcoming" meetings
  • limit (optional): Limit number of results
• Returns: Meeting details with dates, locations, slides, and recordings

query-eboard-members

Query executive board members

• Parameters:
  • memberType (optional): "current", "past", or "all"
  • limit (optional): Limit number of results
• Returns: Member profiles with positions, LinkedIn, GitHub, and photos

query-hackathons

Query hackathon events

• Parameters:
  • status (optional): "ongoing", "upcoming", "past", or "all"
  • slug (optional): Get specific hackathon by slug/ID
  • limit (optional): Limit number of results
• Returns: Hackathon details with dates, registration links, and status

query-graphics

Query landing page graphics

• Parameters:
  • title (optional): Get specific graphic by title
  • limit (optional): Limit number of results
• Returns: Graphic assets with titles, descriptions, and image URLs

query-banners

Query parallax banners

• Parameters:
  • limit (optional): Limit number of results
• Returns: Banner data with titles, links, and image URLs

search-content

Universal search across all content types

• Parameters:
  • query: Search term to match against titles and descriptions
  • contentTypes (optional): Specific content types to search in
  • limit (optional): Limit results per content type (default: 5)
• Returns: Matching content across specified or all content types

Setup

Environment Variables

Create a .env.local file with your Contentful credentials:

CONTENTFUL_SPACE_ID=your_space_id
CONTENTFUL_ACCESS_TOKEN=your_access_token
REDIS_URL=your_redis_url # Required for SSE transport on Vercel

Installation

npm install
# or
yarn install
# or
pnpm install

Development

npm run dev
# or
yarn dev
# or
pnpm dev

Deployment on Vercel

Requirements

• Redis database attached to your Vercel project under REDIS_URL
• Fluid Compute enabled for efficient execution
• Contentful space with the supported content types

Configuration

1. Set up your Contentful environment variables in Vercel
2. Enable Fluid Compute in your Vercel project settings
3. For Pro/Enterprise accounts, adjust maxDuration to 800 in app/[transport]/route.ts
4. Deploy using the Next.js MCP template

Usage Examples

Querying Blog Posts

{
  "tool": "query-blog-posts",
  "parameters": {
    "limit": 5
  }
}

Searching Content

{
  "tool": "search-content",
  "parameters": {
    "query": "hackathon",
    "contentTypes": ["hackathon", "meeting"],
    "limit": 3
  }
}

Getting Current E-board Members

{
  "tool": "query-eboard-members",
  "parameters": {
    "memberType": "current"
  }
}

Testing

Use the included test client to verify your deployment:

node scripts/test-client.mjs https://your-deployment-url.vercel.app

Technical Details

• Framework: Next.js with App Router
• MCP Adapter: @vercel/mcp-adapter
• CMS: Contentful
• Validation: Zod schemas
• Transport: HTTP and SSE (Server-Sent Events)
• Maximum Duration: 60 seconds per request

Content Type Schema

The server exposes detailed schema information through the contentful-schema resource, including field definitions for:

• Rich text content and media assets
• Date/time fields and status enums
• Relationship fields between content types
• Localization and metadata fields

This makes it easy for AI assistants to understand your content structure and generate appropriate queries.