ENSIA

The ENSIA MCP server exposes an API to navigate, search, and analyze your Google Drive-based academic structure organized by cycles, semesters, and modules. It provides tools for hierarchical navigation, global and scoped search, path validation, and module comparison, enabling an AI assistant to understand and reason about your academic materials. With the server running, you can query actions such as listing cycles, listing semesters within a cycle, listing modules in a semester, performing subject-based module discovery, and comparing two modules side-by-side. The server includes dedicated services for search, navigation, and comparison, and models that map the academic structure into searchable artifacts for efficient retrieval by an MCP-enabled assistant like Claude Desktop. Configuration can be customized via environment variables and standard MCP wiring in Claude or other MCP clients.

Prerequisites:

• Python 3.8+
• Access to Google Drive with the expected academic structure layout
• MCP-compatible AI assistant setup (e.g., Claude Desktop)

Installation steps:

1. Clone the repository

git clone <repository-url>
cd MCP-ENSIA

2. Install dependencies

pip install -r requirements.txt

3. (Optional) Run a quick import/test check

# Test imports and basic functionality
python -c "
import sys
sys.path.append('src')
from src.clients.drive_client import EnhancedGoogleDriveClient
from src.core.academic_structure import AcademicStructureProcessor
print('All imports successful!')
" 

4. Start the MCP server

# Start the MCP server
python -m src.server

5. Configure MCP integration (example for Claude Desktop)

{
  "mcpServers": {
    "academic-structure": {
      "command": "python",
      "args": ["-m", "src.server"],
      "cwd": "/path/to/MCP-ENSIA",
      "env": {
        "PYTHONPATH": "/path/to/MCP-ENSIA"
      }
    }
  }
}

Tips:

• Ensure the Google Drive credentials and OAuth token files are present under config/ (credentials.json and token.json) and properly referenced by the server.
• The server relies on Python 3.8+ and the Google Drive API; make sure API access is configured for your environment.
• If you modify the repository layout, update the PYTHONPATH accordingly so Python modules can be discovered.
• For troubleshooting, inspect the tests in tests/ to verify imports and core functionality, and run pytest selectively during development.
• When sharing configurations, avoid exposing sensitive credentials in the mcp configuration payload.