@noteshell/mcp
Access your notes, tasks, and learning plans from Noteshell through this MCP server. Provides 43 pure-data tools for notes, projects, secretary, context, calendar, and search.
README
@noteshell/mcp
Noteshell MCP Server — Access your notes, tasks, and learning plans from Claude Code and any MCP-compatible client. 43 pure-data tools, zero AI calls, 2-minute setup.
What is Noteshell?
Noteshell is an agent-first learning OS for students, researchers, and lifelong learners. Your notes, study plans, daily tasks, and knowledge base — all accessible to your AI assistant through this MCP server.
Setup (2 steps)
1. Authenticate
# Recommended: browser-based login
npx @noteshell/mcp login
# Legacy alternative: email/password
npx @noteshell/mcp setup <your-email> <your-password>
This logs you into Noteshell and saves your credentials to ~/.noteshell.json. You need a Noteshell account.
2. Add to your AI coding tool
Add to your ~/.claude.json, .mcp.json, or equivalent MCP config:
{
"mcpServers": {
"noteshell": {
"command": "npx",
"args": ["-y", "@noteshell/mcp"]
}
}
}
Reload your AI coding tool — done. Works with Claude Code, Cursor, Codex CLI, and any MCP-compatible client.
Available Tools (43 total)
| Prefix | Count | What it does |
|---|---|---|
notes_* |
13 | Create, read, update, search, organize notes |
projects_* |
1 | List your projects/folders |
secretary_* |
20 | Daily tasks, study plans, memory files, analytics |
context_* / soul_* |
5 | Cross-agent context bus, user goals & preferences |
calendar_* |
3 | Calendar events |
search_* |
2 | Full-text search across notes and memory |
Skills
Noteshell ships with 7 methodology skills — step-by-step workflows that teach your AI agent how to use the tools effectively:
| Skill | Purpose |
|---|---|
| Morning Routine | Review yesterday, archive, pick focus, generate plan |
| Evening Reflection | Capture mood, carry over tasks, prepare tomorrow |
| Weekly Review | 7-day analytics, plan progress, mood trends |
| Study Planning | Create roadmaps, break down subjects, track progress |
| Note Organization | Structure projects, enrich notes, task-to-note pattern |
| Research to Notes | Capture research findings as structured knowledge |
| Daily Workflow | Activity logging, quick task updates during the day |
Once the MCP server is running, just talk to your AI agent naturally:
- "Plan my day"
- "What have I been working on?"
- "Create a study plan for algorithms"
- "How am I doing this week?"
How it works
Claude Code / Cursor / Codex CLI
│
▼
@noteshell/mcp (this package — data only, no AI)
│
▼
Your Noteshell data (cloud, secured by your account)
No AI runs inside the MCP server. Claude does the thinking; Noteshell provides the data.
Skill (Methodology Workflows)
For structured workflows (morning routine, evening reflection, weekly review, study planning), install the companion skill:
npx skills add Notecase/noteshell-mcp
This teaches your AI agent how to use the MCP tools effectively — slash commands like /plan-my-day, /morning-routine, /evening-reflection, /weekly-review.
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。