Hjarni MCP Server
Gives AI assistants long-term memory by letting them search, read, create, and organize Markdown notes, folders, tags, and links through a hosted note-taking backend.
README
<p align="center"> <img src="docs/icon.png" alt="Hjarni" width="80" /> </p>
<h1 align="center">Hjarni</h1>
<p align="center"> <strong>Give your AI long-term memory.</strong><br> Note-taking app with a built-in MCP server. </p>
<p align="center"> <a href="https://hjarni.com">Website</a> · <a href="https://hjarni.com/docs">Docs</a> · <a href="https://hjarni.com/blog">Blog</a> </p>
What it does
Hjarni is an AI-native note-taking app. Write notes in Markdown, organize them in folders with custom AI instructions per folder, and let Claude or ChatGPT search, read, create, and organize your notes via MCP.
Your notes serve two readers: you and your AI. Hjarni stores the knowledge. ChatGPT and Claude are the interfaces.
<p align="center"> <img src="docs/claude-memory-demo.gif" alt="Claude searching and creating notes in Hjarni" width="600" /> </p>
Connect
Hjarni is a hosted remote MCP server with OAuth authentication. No API keys, no Docker, no local setup. Setup takes under 2 minutes.
Claude.ai / Claude iOS
- Open Settings > Integrations
- Tap Add custom connector
- Set the name to
hjarniand the URL tohttps://hjarni.com/mcp - Tap Add and log in to Hjarni when redirected
<img src="docs/claude-add-connector.png" alt="Claude Add Connector dialog" width="360" />
Requires Claude Pro or Team plan for MCP integrations.
Claude Desktop / Claude Code
Add to your MCP config (claude_desktop_config.json or .claude.json):
{
"mcpServers": {
"hjarni": {
"url": "https://hjarni.com/mcp",
"type": "streamable-http"
}
}
}
Works on any Claude plan. Authenticates via OAuth on first connection.
ChatGPT
- Go to chatgpt.com > Settings > Connectors
- Enable Developer mode
- Tap Create app
- Set the name to
hjarniand the MCP Server URL tohttps://hjarni.com/mcp - Set authentication to OAuth
- Tap Create and log in to Hjarni when redirected
<img src="docs/chatgpt-new-app.png" alt="ChatGPT Create App form" width="360" />
Requires ChatGPT Plus, Pro, or Team plan.
Other MCP clients
Any client that supports streamable HTTP transport can connect. The endpoint is https://hjarni.com/mcp. Authentication uses OAuth 2.0 with PKCE. Discovery metadata is at /.well-known/oauth-authorization-server.
Available tools
| Tool | Description |
|---|---|
search |
Full-text search across notes, containers, and tags |
notes-create |
Create a new note with title, body, tags, and container placement |
notes-get |
Read a single note including full content, tags, and linked notes |
notes-list |
List notes with filtering by container, tags, and sorting |
notes-update |
Update content, move notes, change tags, archive/unarchive |
notes-delete |
Permanently delete a note |
containers-list |
List folders for organizing notes |
containers-create |
Create new folders |
containers-get |
Get a single container with its AI instructions |
containers-update |
Update a container (rename, move, change description) |
tags-list |
List all tags |
tags-create |
Create new tags |
links-manage |
Create or remove bidirectional links between notes |
instructions-get |
Read AI instructions set on a folder |
instructions-update |
Update folder-level AI instructions |
files-attach |
Attach a file to a note |
files-attach_from_url |
Fetch a file from a URL and attach it to a note |
files-get_download_url |
Get a temporary download URL for a file |
files-remove |
Remove a file attachment |
dashboard-get |
Overview of the account: note count, containers, tags, inbox |
teams-list |
List all teams the user is a member of |
teams-get |
Get team details including recent notes |
Built-in prompts
The server also exposes MCP prompts for clients that support prompt discovery:
| Prompt | Description |
|---|---|
summarize_note |
Summarize a note and suggest tags and related links |
weekly_review |
Review recent activity and suggest organization improvements |
research_topic |
Synthesize everything in the knowledge base related to a topic |
Usage examples
Research assistant
"What decisions have I made about the tech stack?"
Claude searches your notes, follows wiki-links between related notes, and synthesizes a summary from multiple sources.
Quick capture
"Save this as a note in my Projects folder and tag it"
Claude finds the right container, creates the note with Markdown formatting, adds tags, and links it to related notes.
Weekly review
"Help me triage my inbox"
Claude lists unfiled notes, suggests where each belongs based on your folder structure, and moves them with your approval.
Writing with context
"Draft a project update using my recent meeting notes"
Claude pulls relevant notes and uses them as source material.
Features
- Markdown notes with full formatting support
- Wiki-links (
[[id:Note Title]]) for cross-referencing - Folders with custom AI instructions per folder
- Full-text search across all content
- Bidirectional linking between notes
- Tags for cross-cutting categorization
- File attachments (Pro plan and above)
- Team collaboration with shared knowledge bases
- REST API for custom integrations
- Export your data as Markdown at any time
MCP server details
| Endpoint | https://hjarni.com/mcp |
| Transport | Streamable HTTP with JSON-RPC 2.0 |
| Protocol versions | 2025-03-26 and 2024-11-05 |
| Authentication | OAuth 2.0 (PKCE) or Bearer token |
| Capabilities | Tools and prompts |
For the full protocol reference, see the MCP server docs.
Pricing
| Plan | Price | What you get |
|---|---|---|
| Free | $0 | 25 notes, containers, tags, search, full MCP & API access |
| Pro | $10/mo | Unlimited notes, file attachments, share with 2 collaborators |
| Teams | $13/seat/mo | Shared knowledge base, team MCP & API, member management |
Also available in EUR and GBP. See pricing.
Links
- Website
- Docs
- Getting started
- MCP server reference
- REST API reference
- Claude setup guide
- ChatGPT setup guide
- Blog
- Privacy
- Terms
Support
Email evert@hjarni.com.
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。