School MCP
Fetches upcoming assignment deadlines, lists available courses, and downloads course materials and files
Adds deadlines to macOS Calendar or Reminders using AppleScript and manages local file downloads
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@School MCP what assignments are due this week?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
School MCP
A Model Context Protocol (MCP) server for academic tools, integrating with Canvas and Gradescope platforms.
Features
Assignment Deadlines: Fetch and display upcoming deadlines from Canvas and Gradescope
Calendar Integration: Add deadlines to macOS Calendar or Reminders using AppleScript
File Management: Download course materials from Canvas
Related MCP server: moodle-mcp
Quickstart
Installation
Installing via Smithery
To install School MCP for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @54yyyu/school-mcp --client claudeClone the repository:
git clone https://github.com/yourusername/school-mcp.git
cd school-mcpInstall the package:
pip install -e .Set up your environment variables by either:
Using the included setup helper (recommended)
Creating a
.envfile manually
Using the Setup Helper
Run the setup helper to configure Claude Desktop automatically:
python setup_helper.pyThe setup helper will:
Find your Claude Desktop configuration file
Create a
.envfile if neededConfigure the MCP server with proper paths
Add your environment variables to the Claude Desktop configuration
Manual Setup
If you prefer to set up manually:
Copy the environment template:
cp .env.template .env
# Edit .env with your credentialsConfigure Claude Desktop by following the Claude Desktop Integration Guide.
Running the server
Run directly:
python -m school_mcpOr use the convenience script:
./run_server.pyCLI: cw / coursework
pip install -e . also installs a command-line tool, cw (long name coursework):
cw deadlines # unsubmitted work due in the next 14 days (Canvas + Gradescope)
cw deadlines -s canvas -d 7 # Canvas only, next week
cw announcements -u # unread course announcements
cw inbox -u # unread Canvas inbox conversations
cw message 512758 # read a thread (does not mark it read)
cw grades 20.201 # per-assignment scores for one course
cw files 6.7810 --compare ~/Desktop/MIT/6.7810 --missing # what's not filed locally yet
cw page 7.36 L15 # read a lecture page; its files grouped by label (Main Papers, ...)
cw download crispr # download a course's files (or --id ID ... for a selection; -j N files at once)
cw use columbia # switch the active accountCourses can be named by code, name fragment or Canvas id. Every command has
-h, --json, and (where it talks to a school) -a ACCOUNT.
A Claude Code skill for the CLI lives in skill/coursework/; symlink it into
~/.claude/skills/ to make it available in every session.
MCP Tools
get_deadlines: Fetch upcoming assignment deadlines from Canvas and Gradescopeadd_to_reminders: Add assignments to macOS Reminderslist_courses: List all available Canvas courseslist_course_files: List a course's files (with the page label each sits under) without downloadingget_course_page: Read a course Page, or list a course's pagesdownload_course_files: Download files from a Canvas courseset_download_path: Configure where downloaded files are savedget_download_path_info: Check the current download location
Configuration
The server tries to find configuration in this order:
Environment variables
.envfile in the current directoryExisting
config.jsonfile in the home directory
License
MIT
Available Tools
10 toolsadd_to_remindersB
Add upcoming deadlines to macOS Reminders.
Args: days_ahead: Number of days to look ahead for assignments (default: 14) account: School account to use (defaults to the active account)
Inputs: account, days_ahead. Returns: result
download_course_filesA
Download files from a Canvas course (all of them, or a selection).
Args: course_id: Canvas course ID download_path: Root to download into for this call (default: the saved download path; never changes it) file_ids: Only these file ids (from list_course_files) since: Only files created/updated since YYYY-MM-DD or N days ago ("7d") flat: Put files directly in the root instead of // folders account: School account to use (defaults to the active account)
Inputs: course_id (required), flat, since, account, +2 more. Returns: result
get_course_pageA
Read a Canvas Page, such as a lecture page with its readings. Without
page, list the course's pages (module pages first; many courses hide the
Pages list from students).
Args: course: Course id, code (e.g. "7.36") or name fragment page: Piece of the page title (e.g. "L15") or its url slug account: School account to use (defaults to the active account)
Returns the page text plus its file links, each with section (the label it
sits under on the page, e.g. "Main Papers" or "Additional reading") and
hidden (true when the link has no visible text).
Inputs: course (required), page, account. Returns: result
get_deadlinesC
Get upcoming deadlines from Canvas and Gradescope.
Args: days_ahead: Number of days to look ahead for assignments (default: 14) account: School account to use (defaults to the active account)
Inputs: account, days_ahead. Returns: result
get_download_path_infoA
Get the current default download path for Canvas files.
Inputs: no named parameters. Returns: result
list_course_filesA
List every file download_course_files would fetch for a course, without downloading: modules, assignments (attachments + description links), Files tab, Pages, announcements, deduped by file id.
Args: course: Course id, code (e.g. "20.201") or name fragment since: Only files created/updated since this date (YYYY-MM-DD) or N days ago ("7d") compare_dir: Local folder to check against; each file gets local.status have / changed / missing (matched by name+size, then content hash; any layout) account: School account to use (defaults to the active account)
Inputs: course (required), since, account, compare_dir. Returns: result
list_coursesC
List available courses from Canvas.
Args: account: School account to use (defaults to the active account)
Inputs: account. Returns: result
list_school_accountsA
List the configured school accounts (e.g. columbia, mit) and show which one is active.
Inputs: no named parameters. Returns: result
set_download_pathC
Set the default download path for Canvas files.
Args: path: Path to save downloaded files to
Inputs: path (required). Returns: result
switch_school_accountB
Switch the active school account. Persists across restarts.
Args: account: Account name, e.g. "columbia" or "mit"
Inputs: account (required). Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 10 tools
Each tool targets a distinct resource+action: account listing/switching, deadline aggregation, reminders export, course listing, file preview vs. download, page reading, and download-path get/set. The only potentially overlapping pair (list_course_files vs. download_course_files) is explicitly framed as dry-run preview versus actual fetch, and get/set_download_path are a clean complementary pair.
All ten names follow a strict verb_noun snake_case pattern (list_, switch_, get_, add_, download_, set_) with no camelCase or bare-noun outliers. Pluralization and object naming (course_files, school_account, download_path) are used consistently across related tools.
Ten tools is well within the sweet spot and every one earns its place: account management (2), deadlines/reminders (2), courses and files (3), pages (1), path config (2). Nothing looks like filler, and no obvious operation is split into redundant variants.
The surface covers the core student workflow end-to-end: pick an account, list courses, preview and download files, read pages, pull deadlines, and push them to Reminders. Minor gaps remain — no way to add/remove a school account, no per-assignment detail or grades view, and no way to clear/reset the download path — but agents can work around these.
Maintenance
Related MCP Connectors
Model Context Protocol server for Studex tools, notifications, and profile integrations
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
Remote MCP server for notes, tasks, kanban boards and docs - the Slates agentic workspace
Markdown-based note-taking with a hosted MCP server. Your notes serve you and your AI.
Related MCP Servers
- AlicenseBqualityDmaintenanceA Model Context Protocol server that enables interaction with the Canvas Learning Management System API, allowing users to manage courses, assignments, enrollments, and grades within Canvas.543,565 npm104JavaScriptMIT
- FlicenseNot gradedqualityBmaintenanceA Model Context Protocol (MCP) server that connects AI coding agents to your Moodle LMS. Fetch assignments, grades, deadlines, and sync everything to Obsidian automatically.-
- AlicenseNot gradedqualityDmaintenanceMCP server that integrates GitHub, Brightspace, Outlook, and local files, enabling natural language queries for assignments, grades, emails, calendar events, repositories, and file operations.MIT
- AlicenseNot gradedqualityBmaintenanceEnables natural-language queries across multiple schools' Canvas and Blackboard accounts for courses, notifications, assignments, grades, feedback, and schedules, with read-only access and local todo/calendar export.9MIT