easycslearning
Gives Claude read-only access to EasyCS Learning Firebase data, enabling queries about courses, rosters, grades, progress, activity, and integrity flags.
README
EasyCS Learning — MCP Server
Gives Claude read access to your EasyCS Learning Firebase data so you can ask questions like:
"Get 'Yod 4' class grades in lesson 'Variables' in course 'C# for Beginners' and send a WhatsApp message to all students who failed."
Available Tools (17 total)
| Group | Tool | What it does |
|---|---|---|
| Discovery | list_courses |
All active courses |
| Discovery | list_topics |
Topics in a course |
| Discovery | list_lessons |
Lessons in a course / topic |
| Discovery | list_classes |
Classes for a course |
| Roster | get_class_roster |
Students in a class with streak & points |
| Roster | get_student_info |
Full profile for one student |
| Grades | get_lesson_grades |
All grades for a lesson |
| Grades | get_failing_students |
Students below a score threshold |
| Grades | get_student_grades |
All lesson grades for one student |
| Grades | get_grade_matrix |
Full student × lesson grid |
| Progress | get_course_progress |
Completion % per student |
| Progress | get_at_risk_students |
Medium / high risk students |
| Progress | get_student_summary |
Complete snapshot for one student |
| Activity | get_inactive_students |
Not active in N days |
| Activity | get_daily_activity |
Daily active students & answers chart |
| Activity | get_lesson_attendance |
Who submitted vs who didn't |
| Integrity | get_integrity_flags |
Plagiarism / suspicion flags |
One-Time Setup
Step 1 — Create an OAuth2 Client ID
- Go to Google Cloud Console → select project easycslearning-web-app
- Navigate to APIs & Services → Credentials
- Click Create Credentials → OAuth 2.0 Client ID
- Application type: Desktop app
- Name:
EasyCS MCP(or anything) - Click Create → Download JSON
- Create the directory and save the file:
mkdir %USERPROFILE%\.easycs-mcp
Save the downloaded JSON as:
C:\Users\<your-username>\.easycs-mcp\client_secret.json
Step 2 — Enable the Firestore API (if not already enabled)
Go to APIs & Services → Library and enable:
- Cloud Firestore API
Step 3 — Build the server
cd easycslearning_mcp
npm install
npm run build
Step 4 — First login
Run the server once manually to complete the browser login:
node dist/index.js
Your browser will open → sign in with the Google account that owns the Firebase project → the token is saved to ~/.easycs-mcp/credentials.json.
After this, every subsequent start is silent (no browser needed).
Claude Desktop Configuration
Add this to your Claude Desktop config file:
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"easycslearning": {
"command": "node",
"args": [
"C:\\Users\\ronto\\source\\repos\\teaching\\easycslearning_mcp\\dist\\index.js"
]
}
}
}
Restart Claude Desktop. You should see "easycslearning" in the tools list.
Claude Code Configuration
Add to your .claude/settings.local.json (already exists at the repo root):
{
"mcpServers": {
"easycslearning": {
"command": "node",
"args": [
"C:\\Users\\ronto\\source\\repos\\teaching\\easycslearning_mcp\\dist\\index.js"
]
}
}
}
Example Prompts
List all my active courses.
Show me the Yod 4 class roster for "C# for Beginners".
Get grades for lesson "Variables" in "C# for Beginners" for class "Yod 4", then list students who scored below 60%.
Who are the at-risk students in "Python Fundamentals"?
Show me students who haven't been active in the last 10 days in "C# for Beginners".
Get integrity flags with high severity for "C# for Beginners".
Give me a full summary for student "David Cohen" in "C# for Beginners".
Security Notes
- The
client_secret.jsonandcredentials.jsonfiles live in~/.easycs-mcp/on your local machine only — they are never committed to git. - The OAuth2 token is scoped to read-only Firestore access — it cannot write or delete any data.
- This server is designed for local use only (stdio transport). It is not a public HTTP server.
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。