obsidian-mcp-server

obsidian-mcp-server

Connects Claude.ai to your local Obsidian vault for full CRUD access, search, and daily note creation via the Model Context Protocol.

Category
Visit Server

README

obsidian-mcp-server

npm version License: MIT Node.js Version

A production-grade Model Context Protocol (MCP) server that connects Claude.ai to your local Obsidian vault. Gives Claude full CRUD access to your notes, full-text fuzzy search, backlink resolution, frontmatter parsing, and daily note creation — all over the secure stdio transport.


Prerequisites

  • Node.js ≥ 20.0.0
  • npm ≥ 9.0.0
  • An existing Obsidian vault on your local filesystem
  • Claude Desktop (macOS, Windows, or Linux)

Installation

Option A — Clone & build locally

git clone https://github.com/yourusername/obsidian-mcp-server.git
cd obsidian-mcp-server
npm install
npm run build

Option B — Install globally via npm (once published)

npm install -g obsidian-mcp-server

Configure the vault path

cp .env.example .env
# Edit .env and set OBSIDIAN_VAULT_PATH to your vault directory

Or pass it on the CLI:

node dist/index.js --vault /path/to/your/vault

Connecting to Claude Desktop

Add the server to your Claude Desktop config file.

macOS

~/Library/Application Support/Claude/claude_desktop_config.json

Windows

%APPDATA%\Claude\claude_desktop_config.json

Linux

~/.config/claude/claude_desktop_config.json


Config snippet (local build)

{
  "mcpServers": {
    "obsidian": {
      "command": "node",
      "args": ["/absolute/path/to/obsidian-mcp-server/dist/index.js"],
      "env": {
        "OBSIDIAN_VAULT_PATH": "/absolute/path/to/your/obsidian/vault"
      }
    }
  }
}

Config snippet (global npm install)

{
  "mcpServers": {
    "obsidian": {
      "command": "obsidian-mcp",
      "env": {
        "OBSIDIAN_VAULT_PATH": "/absolute/path/to/your/obsidian/vault"
      }
    }
  }
}

After editing the config, restart Claude Desktop. You should see the Obsidian tools available in the tools panel.


Tools Reference

1. read_note

Read a note's body, frontmatter, word count, and last-modified timestamp.

Input:

{
  "path": "Projects/MyNote.md"
}

Output:

{
  "content": "# My Note\n\nMarkdown body without frontmatter...",
  "frontmatter": { "title": "My Note", "tags": ["project"] },
  "wordCount": 142,
  "lastModified": "2024-03-15T10:30:00.000Z"
}

2. write_note

Create or overwrite a note. Set overwrite: true to replace an existing note.

Input:

{
  "path": "Projects/NewNote.md",
  "content": "# New Note\n\nContent here.",
  "frontmatter": { "tags": ["project", "active"] },
  "overwrite": false
}

Output:

{
  "success": true,
  "path": "Projects/NewNote.md",
  "created": true
}

3. append_note

Append content to an existing note, optionally under a specific heading.

Input:

{
  "path": "Daily Notes/2024-03-15.md",
  "content": "- Completed code review",
  "section": "Tasks"
}

Output:

{
  "success": true,
  "newWordCount": 187
}

4. delete_note

Move a note to .trash/ inside the vault (not permanent deletion). Requires confirm: true.

Input:

{
  "path": "Archive/OldNote.md",
  "confirm": true
}

Output:

{
  "success": true,
  "deletedPath": "Archive/OldNote.md"
}

5. list_notes

List notes with optional folder, tag, and recursion filters.

Input:

{
  "folder": "Projects",
  "tag": "active",
  "recursive": true,
  "limit": 50
}

Output:

{
  "notes": [
    {
      "path": "Projects/Alpha.md",
      "title": "Project Alpha",
      "tags": ["project", "active"],
      "lastModified": "2024-03-15T09:00:00.000Z",
      "wordCount": 320
    }
  ]
}

6. search_notes

Full-text fuzzy search with relevance scoring and highlighted excerpts.

Input:

{
  "query": "machine learning neural network",
  "limit": 10,
  "searchIn": ["title", "content"],
  "tag": "research"
}

Output:

{
  "results": [
    {
      "path": "Research/ML-Notes.md",
      "title": "ML Notes",
      "score": 14.2,
      "excerpt": "…backpropagation through **neural network** layers enables **machine learning** models to…",
      "tags": ["research", "ml"]
    }
  ]
}

7. get_backlinks

Find all notes that link to a given note via [[WikiLinks]] or [text](path.md).

Input:

{
  "path": "Concepts/Recursion.md"
}

Output:

{
  "backlinks": [
    {
      "fromPath": "Algorithms/DFS.md",
      "fromTitle": "Depth-First Search",
      "context": "…DFS uses [[Recursion]] as its core mechanism for traversing…"
    }
  ]
}

8. create_daily_note

Create an Obsidian-style daily note at Daily Notes/YYYY-MM-DD.md.

Input:

{
  "date": "2024-03-15",
  "template": "Templates/Daily.md",
  "additionalContent": "## Meeting Agenda\n\n- Sprint planning"
}

Output:

{
  "path": "Daily Notes/2024-03-15.md",
  "alreadyExisted": false
}

Template tokens: {{date}}2024-03-15, {{time}}09:30 AM, {{title}}2024-03-15.


RAG Workflow Example

A typical search → read → write pattern for AI-augmented note-taking:

User: "Summarize all my notes tagged 'meeting' from this week and create a summary note."

Claude:
1. search_notes({ query: "meeting", tag: "meeting", limit: 20 })
   → finds 5 recent meeting notes

2. read_note({ path: "Meetings/2024-03-13.md" })
   read_note({ path: "Meetings/2024-03-14.md" })
   read_note({ path: "Meetings/2024-03-15.md" })
   → reads each note's content

3. write_note({
     path: "Summaries/Week-2024-03-11.md",
     content: "# Week Summary\n\n...",
     frontmatter: { tags: ["summary", "weekly"] }
   })
   → creates the summary note

Security

Path Traversal Protection

Every file path provided to the server is validated against the vault root:

  1. The path is resolved with path.resolve() against the vault root.
  2. The resolved absolute path is checked to ensure it starts with the vault root directory.
  3. Any path that escapes the vault (e.g., ../../etc/passwd, /absolute/paths) throws a VaultSecurityError and is rejected before any I/O occurs.

Atomic Writes

All write operations use a write-to-temp-then-rename pattern:

  1. Content is written to a .tmp.PID.TIMESTAMP file.
  2. The temp file is atomically renamed to the final path.
  3. On failure, the temp file is cleaned up.

This prevents partial writes from corrupting your notes.

Trash Instead of Delete

delete_note moves notes to .trash/ inside the vault rather than permanently deleting them. Notes can be recovered manually from that folder.


Development

# Run in dev mode with hot reload
npm run dev

# Type check
npm run typecheck

# Run tests
npm test

# Run tests in watch mode
npm run test:watch

# Lint
npm run lint

# Build for production
npm run build

Contributing

  1. Fork the repository
  2. Create a feature branch: git checkout -b feat/my-feature
  3. Make your changes with tests
  4. Ensure all tests pass: npm test
  5. Ensure no type errors: npm run typecheck
  6. Submit a pull request

Please follow the existing code style and ensure exactOptionalPropertyTypes strict mode is maintained.


License

MIT License

Copyright (c) 2024 obsidian-mcp-server contributors

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured