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.
README
obsidian-mcp-server
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:
- The path is resolved with
path.resolve()against the vault root. - The resolved absolute path is checked to ensure it starts with the vault root directory.
- Any path that escapes the vault (e.g.,
../../etc/passwd,/absolute/paths) throws aVaultSecurityErrorand is rejected before any I/O occurs.
Atomic Writes
All write operations use a write-to-temp-then-rename pattern:
- Content is written to a
.tmp.PID.TIMESTAMPfile. - The temp file is atomically renamed to the final path.
- 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
- Fork the repository
- Create a feature branch:
git checkout -b feat/my-feature - Make your changes with tests
- Ensure all tests pass:
npm test - Ensure no type errors:
npm run typecheck - 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
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
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.
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.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
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.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.