Coursera MCP

A Model Context Protocol (MCP) server integrating Coursera with Claude AI. Access courses, programs, enrollment data, and personalized recommendations directly within Claude conversations.

Features

7 Powerful Tools:

Public Tools (No Authentication)

• 🔍 search_courses - Search for courses with filters (level, language, sorting)
• 🎯 search_programs - Find specializations, degrees, and certificates
• 📚 get_course_details - Comprehensive course information (syllabus, instructors, skills)
• 🏆 get_program_details - Full program details with course curriculum

Private Tools (TOTP 2FA Required)

• 📖 get_enrolled_courses - List your enrolled courses with progress
• 📊 get_progress - Detailed progress tracking (current week, upcoming deadlines)
• 🎓 get_recommendations - AI-powered course recommendations based on your learning

Installation

Prerequisites

• Node.js >= 18.0.0
• Claude Desktop (for desktop integration)
• Or use directly with Claude API

Via npm

npm install -g coursera-mcp

From Source

git clone https://github.com/fmbetancourt/coursera-mcp.git
cd coursera-mcp
bun install
bun run build
npm link

Quick Start

1. Initial Setup

coursera-mcp init

This will:

• Prompt for your Coursera email
• Display a QR code for 2FA setup (Google Authenticator, Authy, 1Password, etc.)
• Save encrypted session token to ~/.coursera-mcp/sessions.json

2. Use with Claude

In Claude Desktop

Add to your Claude configuration (~/.claude/claude.json):

{
  "mcpServers": {
    "coursera": {
      "command": "coursera-mcp"
    }
  }
}

Restart Claude Desktop → Coursera tools available in conversations

Via Claude API

from anthropic import Anthropic

client = Anthropic()
# MCP servers configured in your environment
response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    tools=[...],  # MCP tools injected
    messages=[{
        "role": "user",
        "content": "Find Python beginner courses"
    }]
)

Usage Examples

Search for Courses

In Claude:

"Find beginner Python courses with at least 4.5 star rating"

Tool Call:

search_courses({
  query: "Python",
  level: "beginner",
  sortBy: "rating",
  sortOrder: "desc"
})

Response:

{
  items: [
    {
      id: "python-basics",
      name: "Python for Everybody",
      rating: 4.8,
      enrollments: 250000,
      duration: 8,
      level: "beginner"
    },
    ...
  ],
  total: 1250,
  hasMore: true
}

Get Course Details

get_course_details({
  courseId: "python-for-data-science"
})

Check Your Progress

get_progress({
  courseId: "machine-learning-specialization"
})

→ {
  percent: 65,
  currentWeek: 7,
  totalWeeks: 10,
  upcomingDeadlines: [
    {
      week: 7,
      dueDate: "2026-05-15T23:59:59Z",
      type: "quiz"
    }
  ]
}

Get Personalized Recommendations

get_recommendations({
  limit: 10,
  includeCourses: true,
  includePrograms: true
})

→ {
  recommendations: [
    {
      name: "Advanced Machine Learning",
      matchScore: 92,
      reason: "You've completed Python and Data Science basics",
      difficulty: "advanced",
      rating: 4.8
    },
    ...
  ]
}

Security

2FA Authentication

• TOTP-based - Industry-standard Time-based One-Time Password (Google Authenticator, Authy, etc.)
• Credentials never stored - Email/password only used during setup
• Encrypted tokens - AES-256-GCM encryption for session storage
• Auto-refresh - Sessions automatically refreshed before expiration

Privacy

• No logging of sensitive data - Tokens, emails, passwords are automatically redacted
• Local session storage - ~/.coursera-mcp/sessions.json (chmod 0o600)
• No data collection - All data stays on your machine

See SECURITY.md for detailed security information.

Documentation

• API.md - Complete tool specifications with examples
• DEVELOPMENT.md - Architecture and development guide
• CONTRIBUTING.md - How to contribute
• CHANGELOG.md - Release history

Status

Implemented (v0.1.0)

• ✅ All 7 tools functional
• ✅ TOTP 2FA authentication
• ✅ AES-256 encryption
• ✅ Circuit breaker resilience
• ✅ Stale-while-revalidate caching
• ✅ 85%+ test coverage
• ✅ GitHub Actions CI/CD

Roadmap (v1.0+)

• 2FA backup codes UI
• Export progress to CSV
• Google Calendar sync
• Course comparison tool
• Study group finder
• Learning path recommendations

See GitHub Issues for roadmap details.

Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines on:

• Code style and standards
• Testing requirements
• Pull request process
• Commit message format

Troubleshooting

"No active session" error

Solution: Run coursera-mcp init to authenticate

Session expired

Solution: The session will auto-refresh automatically. If it fails, run coursera-mcp init again

Certificate errors on macOS

Solution: Ensure you have the latest Bun version: bun upgrade

Performance

• Cache hit: <5ms response time
• Cache miss: 300-1000ms (Coursera API latency)
• Circuit breaker: Auto-recovers in 60 seconds if API down
• Memory: <50MB footprint

Testing

# Run all tests
bun test

# Run with coverage
bun run test:coverage

# Watch mode
bun run dev

License

MIT © 2026 Freddy Betancourt

Support

• 📖 Check documentation
• 🐛 Report issues
• 💬 Start a discussion

Built with ❤️ by the Coursera MCP team

Made for Claude AI and the community.