Limitless MCP Server
Connects AI assistants to personal lifelog data from Limitless AI, enabling users to search and retrieve memories, recordings, and meeting notes. It allows for natural language querying of personal history and provides AI-generated daily summaries within MCP-compatible applications.
README
Limitless MCP Server
Connect AI assistants to your personal memory through Limitless AI's powerful lifelog system. Access your Pendant recordings, meeting notes, and personal insights directly from Claude, Cursor, or any MCP-compatible AI assistant.
<details> <summary>🚀 Quick Start (2 minutes)</summary>
# Install globally
npm install -g @utlyze/limitless-mcp-server
# Or use directly with npx (recommended)
npx @utlyze/limitless-mcp-server
Add to your AI assistant config:
{
"mcpServers": {
"limitless": {
"command": "npx",
"args": ["-y", "@utlyze/limitless-mcp-server"],
"env": {
"LIMITLESS_API_KEY": "your-api-key"
}
}
}
}
</details>
What is Limitless?
Limitless is a personal AI that captures and organizes everything you've seen, said, or heard. Using their Pendant device or desktop app, Limitless creates a searchable memory of your life that you can query and learn from.
Features
🧠 Personal Memory Access
- List and search through all your lifelogs
- Get AI-powered summaries of your days
- Ask questions about your past conversations and meetings
📝 Smart Organization
- Star important memories
- Export memories as markdown
- Filter by date, source, or content
🤖 AI Integration
- Natural language search through your memories
- AI-generated daily summaries
- Context-aware answers about your past
Installation
<details> <summary>📦 Installation Options</summary>
Global Installation (Recommended)
npm install -g @utlyze/limitless-mcp-server
Local Project Installation
npm install @utlyze/limitless-mcp-server
Direct Usage (No Installation)
npx @utlyze/limitless-mcp-server
</details>
Configuration
<details> <summary>🔧 Configuration Steps</summary>
1. Get Your API Key
- Sign up at limitless.ai
- Pair your Pendant or install the desktop app
- Go to Web App → Developer → Create API Key
⚠️ Security Note: Never share your API key or commit it to version control.
2. Configure Your AI Assistant
Claude Desktop (macOS)
Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"limitless": {
"command": "npx",
"args": ["-y", "@utlyze/limitless-mcp-server"],
"env": {
"LIMITLESS_API_KEY": "your-api-key-here"
}
}
}
}
Cursor
Add to .cursor/mcp.json in your project:
{
"mcpServers": {
"limitless": {
"command": "npx",
"args": ["-y", "@utlyze/limitless-mcp-server"],
"env": {
"LIMITLESS_API_KEY": "your-api-key-here"
}
}
}
}
3. Environment Variables
| Variable | Description | Required | Default |
|---|---|---|---|
LIMITLESS_API_KEY |
Your Limitless API key | ✅ Yes | - |
LIMITLESS_BASE_URL |
API base URL | No | https://api.limitless.ai/v1 |
DEBUG |
Enable debug logging | No | false |
</details>
Available Tools
<details> <summary>🛠️ Tool Reference</summary>
limitless_list_lifelogs
List your lifelogs with powerful filtering options.
Parameters:
timezone- Timezone for date filtering (e.g., "America/New_York")date- Filter by specific date (YYYY-MM-DD)startTime- Start time filter (HH:MM)endTime- End time filter (HH:MM)limit- Number of results (max 100)isStarred- Show only starred lifelogsincludeMarkdown- Include full markdown contentdirection- Sort order (asc/desc)
limitless_get_lifelog
Get detailed information about a specific lifelog.
Parameters:
lifelogId- The ID of the lifelogincludeMarkdown- Include markdown contentincludeHeadings- Include content structure
limitless_search_lifelogs
Search through your memories using natural language.
Parameters:
query- Your search querydateFrom- Start date for searchdateTo- End date for searchsource- Filter by source (pendant/desktop/web/mobile)
limitless_get_daily_summary
Get an AI-generated summary of a specific day.
Parameters:
date- Date to summarize (YYYY-MM-DD)timezone- Your timezoneincludeActionItems- Include action itemsincludeKeyTopics- Include main topics
limitless_ask_ai
Ask questions about your memories and get AI-powered answers.
Parameters:
query- Your questiondateFrom- Context start datedateTo- Context end dateincludeContext- Include source references
limitless_star_lifelog
Mark important memories by starring them.
Parameters:
lifelogId- The lifelog to star/unstarisStarred- true to star, false to unstar
limitless_export_markdown
Export your memories as formatted markdown.
Parameters:
date- Export specific datedateFrom/dateTo- Export date rangeincludeMetadata- Include timestamps and metadatagroupByDate- Organize by date
</details>
Usage Examples
<details> <summary>💡 Natural Language Examples</summary>
Basic Queries
"Show me all my memories from yesterday"
"What did I discuss in meetings today?"
"Find conversations about the product launch"
"Get my daily summary for Monday"
Advanced Searches
"Search for discussions about AI in the last week"
"Show me starred memories from December"
"Export all my meeting notes from this month"
"What action items came up in yesterday's standup?"
AI Questions
"What did John say about the budget proposal?"
"When did I last talk about vacation plans?"
"Summarize my conversations about the new feature"
"What decisions were made in the leadership meeting?"
Organization
"Star the memory about the product roadmap"
"Export this week's meetings as markdown"
"Show me all my Pendant recordings from home"
</details>
Architecture
<details> <summary>🏗️ Technical Details</summary>
Technology Stack
- Language: TypeScript
- MCP SDK: @modelcontextprotocol/sdk
- HTTP Client: Axios
- Validation: Zod
- Node Version: 16+
Project Structure
src/
├── config/ # Configuration management
├── lib/ # Core libraries
│ ├── limitless-client.ts
│ └── logger.ts
├── tools/ # MCP tool implementations
│ └── lifelogs/ # Lifelog-related tools
└── types/ # TypeScript type definitions
Error Handling
- Comprehensive error messages
- Graceful API failure handling
- Detailed debug logging
- Input validation
Security
- API key validation at startup
- Secure credential handling
- No credential logging
- HTTPS-only communication
</details>
Development
<details> <summary>👩💻 Development Guide</summary>
Setup
# Clone the repository
git clone https://github.com/Utlyze/limitless-mcp-server.git
cd limitless-mcp-server
# Install dependencies
npm install
# Build the project
npm run build
# Run tests
npm test
Scripts
npm run dev- Development mode with hot reloadnpm run build- Build for productionnpm test- Run test suitenpm run lint- Check code stylenpm run typecheck- Verify TypeScript types
Testing
# Run all tests
npm test
# Run with coverage
npm run test:coverage
# Run specific test suite
npm run test:unit
npm run test:integration
Contributing
- Fork the repository
- Create a feature branch
- Make your changes
- Add tests
- Submit a pull request
</details>
API Limitations
<details> <summary>⚠️ Current Limitations</summary>
- Pendant Only: Currently limited to Pendant recordings (desktop/web coming soon)
- Read Only: No ability to create or modify lifelogs
- Beta API: Some features may change as the API evolves
- Rate Limits: Standard API rate limits apply
</details>
Troubleshooting
<details> <summary>🔍 Common Issues</summary>
"API Key Required" Error
- Ensure
LIMITLESS_API_KEYis set in your config - Check that the key is valid and active
- Verify you've created a key at app.limitless.ai
"No Lifelogs Found"
- Ensure you have recordings in your account
- Check the date range you're querying
- Verify your Pendant is syncing properly
Connection Issues
- Check your internet connection
- Verify the API is accessible
- Try the debug mode:
DEBUG=true
Tool Not Found
- Restart your AI assistant
- Check the tool name is correct
- Verify MCP server is running
</details>
Support
- Documentation: limitless.ai/developers
- Slack Community: Join #developers channel
- GitHub Issues: Report bugs here
- Email: support@limitless.ai
License
MIT License - see LICENSE file for details.
Built with ❤️ for the Limitless community. Remember everything, forget nothing.
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。