Skylight MCP Server
Enables AI assistants to interact with the Skylight family calendar API, allowing management of calendars, chores, lists, tasks, family members, and rewards through natural language.
README
Skylight MCP Server
An MCP (Model Context Protocol) server for the Skylight Calendar API. Enables AI assistants like Claude to interact with your Skylight family calendar, chores, lists, and more.
Features
- Calendar: Query calendar events ("What's on my calendar today?")
- Chores: View and create chores ("Add emptying dishwasher to chores")
- Lists: View grocery and to-do lists ("What's on the grocery list?")
- Tasks: Add items to the task box ("Add XYZ to my task list")
- Family: View family members and devices
- Rewards: Check reward points and available rewards
Quick Start
Installation
Option 1: npm package (Recommended)
mcp.json:
{
"mcpServers": {
"skylight": {
"command": "npx",
"args": ["@eaglebyte/skylight-mcp"],
"env": {
"SKYLIGHT_EMAIL": "your_email@example.com",
"SKYLIGHT_PASSWORD": "your_password",
"SKYLIGHT_FRAME_ID": "your_frame_id"
}
}
}
}
Claude Code:
claude mcp add skylight npx @eaglebyte/skylight-mcp \
-e SKYLIGHT_EMAIL=your_email@example.com \
-e SKYLIGHT_PASSWORD=your_password \
-e SKYLIGHT_FRAME_ID=your_frame_id
Option 2: From source
git clone https://github.com/TheEagleByte/skylight-mcp.git
cd skylight-mcp && npm install && npm run build
Then use in mcp.json:
{
"mcpServers": {
"skylight": {
"command": "node",
"args": ["/path/to/skylight-mcp/dist/index.js"],
"env": {
"SKYLIGHT_EMAIL": "your_email@example.com",
"SKYLIGHT_PASSWORD": "your_password",
"SKYLIGHT_FRAME_ID": "your_frame_id"
}
}
}
}
Instructions for AI
Copy this into your AI's custom instructions or system prompt:
You have access to the Skylight MCP server. Skylight is a smart family calendar display that shows calendars, chores, grocery lists, meals, and rewards. Use the Skylight tools to help manage family schedules and organization.
Tips:
- Call
get_family_membersbefore assigning chores to get member names- Grocery items default to the main grocery list if no list specified
- Dates accept "today", "tomorrow", day names, or YYYY-MM-DD format
- Some tools (rewards, meals, photos) require Skylight Plus subscription
Prerequisites
- Node.js 18+
- A Skylight account with an active subscription
- Your Skylight Frame ID (see Finding your Frame ID)
Authentication
The MCP server supports two authentication methods:
Option 1: Email/Password (Recommended)
Use your Skylight account credentials. The server will automatically log in and manage tokens.
SKYLIGHT_EMAIL=your_email@example.com
SKYLIGHT_PASSWORD=your_password
SKYLIGHT_FRAME_ID=your_frame_id
Option 2: Manual Token (Legacy)
Capture a token from the Skylight app using a proxy tool.
SKYLIGHT_TOKEN=your_token_here
SKYLIGHT_FRAME_ID=your_frame_id
SKYLIGHT_AUTH_TYPE=bearer
Finding your Frame ID
You still need to find your frame ID (the household identifier):
- Use a proxy tool (Proxyman, Charles, or mitmproxy)
- Capture any API request from the Skylight app
- Look at the URL path:
/api/frames/{frameId}/... - Example:
/api/frames/abc123/chores→ frame ID isabc123
Configuration
| Variable | Required | Description |
|---|---|---|
SKYLIGHT_EMAIL |
Option 1 | Your Skylight account email |
SKYLIGHT_PASSWORD |
Option 1 | Your Skylight account password |
SKYLIGHT_TOKEN |
Option 2 | Your API token (if not using email/password) |
SKYLIGHT_AUTH_TYPE |
No | bearer (default) or basic (for manual token) |
SKYLIGHT_FRAME_ID |
Yes | Your household frame ID |
SKYLIGHT_TIMEZONE |
No | Default timezone (default: America/New_York) |
Example .env file:
# Email/password auth (recommended)
SKYLIGHT_EMAIL=your_email@example.com
SKYLIGHT_PASSWORD=your_password
SKYLIGHT_FRAME_ID=your_frame_id
SKYLIGHT_TIMEZONE=America/New_York
Available Tools
Calendar Tools
| Tool | Description |
|---|---|
get_calendar_events |
Get calendar events for a date range |
get_source_calendars |
List connected calendar sources (Google, iCloud, etc.) |
Chore Tools
| Tool | Description |
|---|---|
get_chores |
Get chores with optional filters (date, assignee, status) |
create_chore |
Create a new chore with optional recurrence |
List Tools
| Tool | Description |
|---|---|
get_lists |
Get all available lists |
get_list_items |
Get items from a specific list |
Task Tools
| Tool | Description |
|---|---|
create_task |
Add a task to the task box |
Family Tools
| Tool | Description |
|---|---|
get_family_members |
Get family member profiles |
get_frame_info |
Get household/frame information |
get_devices |
List Skylight devices |
Reward Tools
| Tool | Description |
|---|---|
get_rewards |
Get available rewards |
get_reward_points |
Get reward points balance |
Example Queries
Once configured, you can ask Claude things like:
- "What's on my calendar today?"
- "What chores do I need to do this week?"
- "Add 'take out trash' to my chores for tomorrow"
- "What's on the grocery list?"
- "Add milk to my task list"
- "Who are the family members on Skylight?"
- "How many reward points does each person have?"
Development
# Run in development mode (with hot reload)
npm run dev
# Build
npm run build
# Run tests
npm test
# Type check
npm run typecheck
API Documentation
This MCP server is built on top of the reverse-engineered Skylight API. The API endpoints were documented using the skylight-api project, which converts browser network traffic (HAR files) into an OpenAPI specification.
API Resources:
If you discover new API endpoints or find issues with the current documentation, please contribute to the skylight-api repository.
Contributing
Contributions are welcome! Here's how you can help:
- Fork the repository and create a feature branch
- Make your changes with clear, descriptive commits
- Run tests (
npm test) and linting (npm run lint) before submitting - Open a pull request with a description of your changes
Development Setup
git clone https://github.com/TheEagleByte/skylight-mcp.git
cd skylight-mcp
npm install
npm run dev # Start with hot reload
Areas for Contribution
- Adding support for new Skylight API endpoints
- Improving error handling and edge cases
- Enhancing documentation
- Writing additional tests
Issues & Support
- Bug reports: Open an issue with steps to reproduce
- Feature requests: Open an issue describing the use case
- Questions: Start a discussion or open an issue
Please include relevant details like your Node.js version, error messages, and configuration (with sensitive values redacted).
License
MIT
Disclaimer
This is an unofficial integration. The Skylight API is reverse-engineered and may change without notice. Use at your own risk.
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。