Eclass MCP Server

🚀 eClass MCP Server

eClass MCP Server provides an independent interface for interacting with the eClass learning management system, integrating AI agents securely and efficiently.

🚀 Quick Start

The eClass MCP Server is an interface independent of the official eClass platform, developed and maintained by the Greek University Network (GUnet) for Greek academic institutions. It adheres to the platform's terms of use and service.

✨ Features

• AI Integration: Supports direct interaction with AI agents (such as Claude or other LLMs) that follow the Model Context Protocol (MCP).
• Course Information: Enables retrieval of course details, viewing of announcements, access to materials, and future support for assignment submission.
• Security Design: Ensures secure handling of credentials to prevent leakage of sensitive data.

📦 Installation

This part is not provided in the original document, so it is skipped.

💻 Usage Examples

Basic Usage

import asyncio
from eclass_mcp_client import MCPClient, ServerConfig

async def main():
    # Configure server parameters
    server_config = ServerConfig(
        host="your_e_class_url",
        username="your_username",
        password="your_password"
    )

    async with MCPClient(server_config) as client:
        try:
            # Login
            login_result = await client.call_tool("login", {"random_string": "dummy"})
            print(login_result)

            # Get course list
            courses_result = await client.call_tool("get_courses", {"random_string": "dummy"})
            print(courses_result)

            # Logout
            logout_result = await client.call_tool("logout", {"random_string": "dummy"})
            print(logout_result)
        except Exception as e:
            print(f"Error occurred: {e}")

if __name__ == "__main__":
    asyncio.run(main())

📚 Documentation

Core Components

1. MCP Server

   • Runs locally.
   • Does not rely on any cloud services.
   • Stores credentials via environment variables.
   • Uses random strings as placeholder parameters.

2. AI Client Isolation

   • AI clients cannot access real credentials.
   • Prevents credentials from being stored or processed on remote servers.
   • Ensures all operations communicate directly with eClass.

Core Protocols

• Stores configuration information using environment variables: e_class_host, e_class_username, e_class_password.
• Supported tool commands:
  • login: Initializes a session and obtains a CAS ticket.
  • get_courses: Returns a list of all courses for the currently logged - in user.
  • logout: Logs out the current session.

Security Mechanisms

• Local Execution: The server runs locally without using any cloud services.
• In - memory Session Management: Session cookies and states are only stored in memory and not persisted to disk.
• Direct Authentication: Communicates directly with eClass through the university's CAS mechanism.

🔧 Technical Details

Credential Protection

• Environment Variable Storage: Credentials are only stored in the local .env file.
• Avoid Plain - text Transmission: All tools use random strings as placeholders, preventing AI clients from accessing real credentials.
• Physical Machine Isolation: Runs only on trusted devices.

Recommended Practices

• Do not commit the .env file to version control.
• Clean up unused server instances in a timely manner.
• Use a strong password policy to protect user accounts.

📄 License

This project is licensed under the GNU General Public License v3.0 (GPLv3), which allows free use, modification, and distribution, but requires the inclusion of the original copyright notice in redistributions or derivative works.