Smartschool Mcp

What is Smartschool Mcp?

Smartschool Mcp is a Model Context Protocol (MCP) server that gives AI assistants secure, dynamically discovered access to Smartschool APIs. It automatically detects all available Smartschool methods and exposes them as MCP tools with built‑in safety guardrails.

How to use Smartschool Mcp?

Install via npm (npm install and build), set environment variables (SMARTSCHOOL_API_ENDPOINT, SMARTSCHOOL_ACCESS_CODE, and optionally ALLOW_DESTRUCTIVE and REQUIRE_CONFIRMATION), then configure your MCP client (e.g., Claude Desktop) to launch the server using npx -y @abrianto/smartschool-mcp or manually via node build/mod.js. The server communicates over stdio.

Key features of Smartschool Mcp

• Dynamic auto-discovery of all Smartschool API methods
• Enterprise safety with four protection levels (SAFE, MODERATE, DESTRUCTIVE, CRITICAL)
• AI-optimized context including Belgian school conventions and absence codes
• Zero maintenance – no manual tool definitions required
• Future-proof – adapts automatically when Smartschool SDK updates
• Built-in domain expertise for Belgian school systems

Use cases of Smartschool Mcp

• Manage users (create, update, delete) with confirmation for destructive actions
• Handle class and group assignments (save, delete classes, assign students)
• Send messages to students and parents via Smartschool’s messaging system
• Check attendance records and academic history
• Perform system administration tasks like synchronisation and support tickets

FAQ from Smartschool Mcp

What are the safety levels and how do they work?

The server classifies operations into SAFE (read‑only, no restrictions), MODERATE (reversible changes, always allowed), DESTRUCTIVE (high‑impact changes, requires ALLOW_DESTRUCTIVE=true), and CRITICAL (permanent deletions, requires ALLOW_DESTRUCTIVE=true and the confirmDestructiveAction parameter).

What are the runtime requirements?

Node.js 18+ or a compatible JavaScript runtime, access to a Smartschool instance, valid API credentials (endpoint and access code), and network connectivity to the Smartschool API endpoint.

How do I enable destructive operations?

Set the environment variable ALLOW_DESTRUCTIVE=true and include confirmDestructiveAction: true in the tool call parameters. By default, destructive operations are blocked in production.

What data does the server access?

The server connects to Smartschool’s V3 API and requires permissions for user management, group/class management, messaging, and reporting. Credentials should be kept secure and not committed to version control.

How does the server handle authentication and transport?

It uses stdio transport per the MCP specification. Authentication is provided via the SMARTSCHOOL_ACCESS_CODE environment variable. No other transports or authentication methods are documented.