SBU Syllabus MCP
An MCP server that transforms course syllabi into structured, queryable data — accessible directly from Claude.ai.
README
SBU Syllabus MCP
An MCP server that transforms course syllabi into structured, queryable data — accessible directly from Claude.ai.
Problem
Students manually dig through multiple syllabi every semester to track exams, deadlines, and policies. This is repetitive and doesn't integrate with AI tools.
Solution
Upload a syllabus PDF to Claude.ai → Claude parses it → stores structured data in MCP server → query anytime with natural language.
┌──────────┐ ┌───────────┐ ┌────────────┐
│ Upload │────▶│ Claude.ai │────▶│ MCP Server │
│ PDF │ │ (parses) │ │ (stores) │
└──────────┘ └───────────┘ └────────────┘
│
▼
"When's my exam?"
│
▼
"May 15, 2:15 PM"
Tech Stack
- Protocol: Model Context Protocol (MCP)
- Server: FastMCP + Starlette + SSE
- Storage: JSON
- Hosting: Railway
- Runtime: Python 3.10+ / uv
Usage
1. Connect
Add to Claude.ai → Settings → Integrations:
https://sbu-syllabus-mcp-production.up.railway.app/sse
2. Upload
Attach PDF and say:
"Parse and upload this syllabus"
3. Query
Exams
"When are my CSE 351 exams?"
"What's the date of the CSE 351 final?"
"Show me all my upcoming exams"
Grading
"How is CSE 351 graded?"
"What percentage is the final worth in LIN 200?"
"What do I need to get an A in CSE 300?"
Policies
"What's the late policy for CSE 351?"
"How many absences am I allowed in CSE 300?"
"Can I make up a missed exam in LIN 200?"
Schedule
"What's the CSE 351 schedule?"
"What topic is covered in week 5 of LIN 200?"
"When do we learn about machine learning in CSE 351?"
Calendar
"Add CSE 351 exams to my Google Calendar"
"Export LIN 200 exam dates to calendar"
"I want to add all my exams to my calendar"
General
"What courses do I have stored?"
"Show me everything about CSE 351"
"Who's the instructor for LIN 200?"
Features
- Syllabus Parsing: Extracts course info, exams, grading, policies, and weekly schedule
- Natural Language Queries: Ask questions about any stored course
- Calendar Export: Export exam dates as ICS file for Google Calendar / Apple Calendar
- Multi-Course Support: Store and query multiple syllabi
- Zero API Cost: Claude.ai handles parsing — no additional API keys needed
- Persistent Storage: Data persists across sessions
Architecture
server.py # MCP server + tools
storage.py # JSON persistence
schema.py # Pydantic models
Data Schema
Syllabus
├── course: { code, title, credits, semester, modality }
├── instructors: [{ name, email, office_hours }]
├── meetings: [{ days, start_time, end_time, location }]
├── grading: [{ name, weight, details }]
├── grade_scale: [{ letter, min_percent, max_percent }]
├── exams: [{ name, date, time, location }]
├── policies: { late_work, attendance, academic_integrity, makeup_exam }
├── schedule: [{ week, topic, due }]
└── textbook
Tools
| Tool | Description |
|---|---|
upload_syllabus |
Store parsed syllabus data |
get_syllabus |
Get full syllabus |
get_exam_dates |
Get exam dates |
get_grading_breakdown |
Get grading weights |
get_policies |
Get course policies |
get_schedule |
Get weekly schedule |
export_calendar |
Export exams as ICS for calendar import |
list_courses |
List stored courses |
Why MCP?
- Native Claude.ai integration — no custom UI needed
- Standardized tool protocol
- SSE streaming built-in
- Works with Claude Desktop too
Local Development
git clone https://github.com/mkjun2016/sbu-syllabus-mcp
cd sbu-syllabus-mcp
uv sync
uv run python server.py
Claude Desktop Config
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"sbu-syllabus": {
"command": "uv",
"args": ["run", "--directory", "/path/to/sbu-syllabus-mcp", "python", "server.py"],
"cwd": "/path/to/sbu-syllabus-mcp"
}
}
}
Deployment
Deployed on Railway with automatic GitHub integration.
# Procfile
web: python server.py
Roadmap
- [x] MCP server setup + schema design
- [x] SSE transport for remote access
- [x] Railway deployment
- [x] Claude.ai integration
- [x] Structured syllabus upload
- [x] Query tools (exams, grading, policies, schedule)
- [x] Calendar export (ICS)
- [ ] Cross-course conflict detection
- [ ] Deadline reminders
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。