Enhanced Obsidian MCP Server with Claude.ai Remote Integration, Tailscale Support, and Advanced Query Capabilities!
π₯ Enhanced Fork Notice: This is an enhanced version of the excellent cyanheads/obsidian-mcp-server with additional features specifically tailored for remote Claude.ai integration, advanced task querying, and security via Tailscale.
An MCP (Model Context Protocol) server providing comprehensive access to your Obsidian vault. Enables LLMs and AI agents to read, write, search, and manage your notes and files through the Obsidian Local REST API plugin.
Built on the cyanheads/mcp-ts-template, this server follows a modular architecture with robust error handling, logging, and security features.
Simultaneous access to multiple Obsidian vaults through a single MCP server:
- Multiple Vault Management: Connect to multiple Obsidian instances on different ports simultaneously
- Vault-Specific Routing: Tools automatically route to the correct vault based on
vaultparameter - Individual Authentication: Separate API keys for each vault with centralized MCP authentication
- Backwards Compatible: Existing single-vault configurations continue to work unchanged
- Dynamic Configuration: JSON-based vault configuration with validation and error handling
Perfect integration with Claude.ai's Remote MCP feature:
- Stateless HTTP Mode: Dedicated stateless transport for Claude.ai compatibility (
MCP_HTTP_STATELESS=true) - Session-Based Mode: Traditional session management for other MCP clients
- Simplified Authentication: Uses dedicated MCP_AUTH_KEY for server access
- Zero Configuration: Works out-of-the-box with Claude.ai Remote MCP servers
- Production Ready: Enterprise-grade stability and error handling
Access your Obsidian vault securely from anywhere:
- Tailscale Funnel Integration: Secure HTTPS endpoints with automatic certificates
- End-to-End Encryption: All traffic encrypted through Tailscale network
- No Port Forwarding: Zero network configuration required
- Access Control: Built-in Tailscale ACL support for enterprise security
Advanced querying capabilities beyond the original:
- Tasks Plugin Integration: Deep integration with Obsidian Tasks plugin
- Advanced Date Parsing: Natural language date recognition
- Priority Detection: Visual and text-based priority parsing
- Multiple Output Formats: Table, list, and summary views
Enterprise-grade monitoring and auto-restart capabilities:
- Health Check Script: Comprehensive component validation (
scripts/health-check.sh) - Intelligent Monitoring: Auto-restart with process lifecycle management (
scripts/monitor-mcp.sh) - macOS Auto-Start: Launch agent configuration for system startup (
scripts/setup-autostart.sh) - Dynamic Port Management: Automatic port conflict resolution (3010-3013 range)
- Enhanced Logging: Detailed connection debugging and API key validation
This server equips your AI with specialized tools to interact with your Obsidian vault:
| Tool Name | Description | Key Features |
|---|---|---|
obsidian_read_file |
Retrieves the content and metadata of a specified file. | - Read in markdown or json format.- Case-insensitive path fallback. - Includes file stats (creation/modification time). |
obsidian_update_file |
Modifies notes using whole-file operations. | - append, prepend, or overwrite content.- Can create files if they don't exist. - Targets files by path, active note, or periodic note. |
obsidian_search_replace |
Performs search-and-replace operations within a target note. | - Supports string or regex search. - Options for case sensitivity, whole word, and replacing all occurrences. |
obsidian_global_search |
Performs a search across the entire vault. | - Text or regex search. - Filter by path and modification date. - Paginated results. |
obsidian_list_files |
Lists files and subdirectories within a specified vault folder. | - Filter by file extension or name regex. - Provides a formatted tree view of the directory. |
obsidian_manage_frontmatter |
Atomically manages a note's YAML frontmatter. | - get, set, or delete frontmatter keys.- Avoids rewriting the entire file for metadata changes. |
obsidian_manage_tags |
Adds, removes, or lists tags for a note. | - Manages tags in both YAML frontmatter and inline content. |
obsidian_delete_file |
Permanently deletes a specified file from the vault. | - Case-insensitive path fallback for safety. |
obsidian_dataview_query |
Execute Dataview DQL queries against your vault. | - Run TABLE, LIST queries using Dataview syntax. - Query notes by tags, frontmatter, dates. - Generate reports and analytics. |
obsidian_task_query |
Search and analyze tasks across your vault. | - Filter by status, date ranges, priorities. - Multiple output formats. - Extract task metadata (due dates, tags). |
| Overview | Features | Installation | | Configuration | Project Structure | Vault Cache Service | | Tools | Resources | Development | License |
The Obsidian MCP Server acts as a bridge, allowing applications (MCP Clients) that understand the Model Context Protocol (MCP) β like advanced AI assistants (LLMs), IDE extensions, or custom scripts β to interact directly and safely with your Obsidian vault.
Instead of complex scripting or manual interaction, your tools can leverage this server to:
- Automate vault management: Read notes, update content, manage frontmatter and tags, search across files, list directories, and delete files programmatically.
- Integrate Obsidian into AI workflows: Enable LLMs to access and modify your knowledge base as part of their research, writing, or coding tasks.
- Build custom Obsidian tools: Create external applications that interact with your vault data in novel ways.
Built on the robust mcp-ts-template, this server provides a standardized, secure, and efficient way to expose Obsidian functionality via the MCP standard. It achieves this by communicating with the powerful Obsidian Local REST API plugin running inside your vault.
Developer Note: This repository includes a .clinerules file that serves as a developer cheat sheet for your LLM coding agent with quick reference for the codebase patterns, file locations, and code snippets.
Leverages the robust utilities provided by the mcp-ts-template:
- Logging: Structured, configurable logging (file rotation, console, MCP notifications) with sensitive data redaction.
- Error Handling: Centralized error processing, standardized error types (
McpError), and automatic logging. - Configuration: Environment variable loading (
dotenv) with comprehensive validation. - Input Validation/Sanitization: Uses
zodfor schema validation and custom sanitization logic. - Request Context: Tracking and correlation of operations via unique request IDs.
- Type Safety: Strong typing enforced by TypeScript and Zod schemas.
- HTTP Transport Option: Native Node.js HTTP server with session management, CORS support, and API key authentication.
- Obsidian Local REST API Integration: Communicates directly with the Obsidian Local REST API plugin via HTTP requests managed by the
ObsidianRestApiService. - Comprehensive Command Coverage: Exposes key vault operations as MCP tools (see Tools section).
- Vault Interaction: Supports reading, updating (append, prepend, overwrite), searching (global text/regex, search/replace), listing, deleting, and managing frontmatter and tags.
- Targeting Flexibility: Tools can target files by path, the currently active file in Obsidian, or periodic notes (daily, weekly, etc.).
- Vault Cache Service: An intelligent in-memory cache that improves performance and resilience. It caches vault content, provides a fallback for the global search tool if the live API fails, and periodically refreshes to stay in sync.
- Safety Features: Case-insensitive path fallbacks for file operations, clear distinction between modification types (append, overwrite, etc.).
- Obsidian: You need Obsidian installed.
- Obsidian Local REST API Plugin: Install and enable the Obsidian Local REST API plugin within your Obsidian vault.
- API Key: Configure an API key within the Local REST API plugin settings in Obsidian. You will need this key to configure the server.
- Node.js & npm: Ensure you have Node.js (v18 or later recommended) and npm installed.
- Tailscale (for remote access): Install Tailscale and enable Tailscale Funnel for secure remote Claude.ai integration.
π‘ Quick Setup: For automatic startup on boot, see the Auto-Start Setup Guide after installation.
- Clone this enhanced repository:
git clone https://github.com/BoweyLou/obsidian-mcp-server-enhanced.git cd obsidian-mcp-server-enhanced - Install dependencies:
npm install
- Build the project:
This compiles the TypeScript code to JavaScript in the
npm run build
dist/directory and makes the entry point executable.
Configure the server using environment variables.
These variables must be set in the MCP client configuration (e.g., cline_mcp_settings.json) or in your environment before starting the server (if running directly).
If running directly, they can be set in a .env file in the project root or directly in your environment.
| Variable | Description | Required | Default |
|---|---|---|---|
MCP_AUTH_KEY |
Authentication key for Claude.ai Remote MCP access. Generate with openssl rand -hex 32 |
Yes (Remote) | undefined |
OBSIDIAN_VAULTS |
JSON array of vault configurations for multi-vault mode. | Yes (Multi) | undefined |
OBSIDIAN_API_KEY |
API Key from Obsidian plugin (single-vault mode only). | Yes (Single) | undefined |
OBSIDIAN_BASE_URL |
Base URL of Obsidian API (single-vault mode only). | Yes (Single) | http://127.0.0.1:27123 |
MCP_TRANSPORT_TYPE |
Server transport: stdio or http. |
No | http |
MCP_HTTP_PORT |
Port for the HTTP server. | No | 3010 |
MCP_HTTP_HOST |
Host for the HTTP server. | No | 127.0.0.1 |
MCP_HTTP_STATELESS |
Enable stateless mode for Claude.ai compatibility. | No | false |
MCP_ALLOWED_ORIGINS |
Comma-separated origins for CORS. Set for production. | No | (none) |
CHATGPT_LAYER_ENABLED |
Set to true to serve the ChatGPT manifest plus JSON action endpoint. |
No | false |
CHATGPT_MANIFEST_PATH |
HTTP path that exposes the ChatGPT manifest JSON. | No | /.well-known/obsidian-chatgpt-manifest.json |
CHATGPT_ACTIONS_PATH |
HTTP path for ChatGPT JSON actions (POST). | No | /chatgpt/actions |
MCP_LOG_LEVEL |
Logging level (debug, info, error, etc.). |
No | info |
OBSIDIAN_VERIFY_SSL |
Set to false to disable SSL verification. |
No | true |
OBSIDIAN_ENABLE_CACHE |
Set to true to enable the in-memory vault cache. |
No | true |
OBSIDIAN_CACHE_REFRESH_INTERVAL_MIN |
Refresh interval for the vault cache in minutes. | No | 10 |
The server supports both single-vault (backwards compatible) and multi-vault modes:
# .env file
MCP_AUTH_KEY=your-generated-mcp-auth-key
OBSIDIAN_API_KEY=your-obsidian-plugin-api-key
OBSIDIAN_BASE_URL=http://127.0.0.1:27123
MCP_TRANSPORT_TYPE=http
MCP_HTTP_STATELESS=true# .env file
MCP_AUTH_KEY=your-generated-mcp-auth-key
OBSIDIAN_VAULTS='[
{
"id": "work",
"name": "Work Vault",
"apiKey": "work-vault-api-key",
"baseUrl": "http://127.0.0.1:27123",
"verifySsl": false
},
{
"id": "personal",
"name": "Personal Vault",
"apiKey": "personal-vault-api-key",
"baseUrl": "http://127.0.0.1:27122",
"verifySsl": false
}
]'
MCP_TRANSPORT_TYPE=http
MCP_HTTP_STATELESS=true- Generate MCP Auth Key:
openssl rand -hex 32 - Configure Multiple Obsidian Instances: Install Local REST API plugin on different ports
- Get API Keys: Extract API keys from each Obsidian instance's plugin settings
- Configure Vaults: Update
.envwithOBSIDIAN_VAULTSJSON configuration - Start Server:
npm run start:http - Access via Claude.ai: Use your Tailscale URL with MCP_AUTH_KEY
To connect in single-vault mode, configure the base URL (OBSIDIAN_BASE_URL) and API key (OBSIDIAN_API_KEY). The Obsidian Local REST API plugin offers two connection types:
-
Encrypted (HTTPS):
- Uses secure
https://endpoint (e.g.,https://127.0.0.1:27124) - Requires
OBSIDIAN_VERIFY_SSL=falsefor self-signed certificates
- Uses secure
-
Non-encrypted (HTTP) - Recommended:
- Uses
http://endpoint (e.g.,http://127.0.0.1:27123) - Simpler setup, no SSL verification needed
- Uses
For multi-vault mode, configure each vault individually in the OBSIDIAN_VAULTS JSON array with its own API key and base URL. Each vault can use either HTTP or HTTPS as needed.
Example configurations:
Single-vault with HTTP:
"env": {
"MCP_AUTH_KEY": "your-generated-mcp-auth-key",
"OBSIDIAN_API_KEY": "your-obsidian-api-key",
"OBSIDIAN_BASE_URL": "http://127.0.0.1:27123"
}Multi-vault configuration:
"env": {
"MCP_AUTH_KEY": "your-generated-mcp-auth-key",
"OBSIDIAN_VAULTS": "[{\"id\":\"work\",\"name\":\"Work\",\"apiKey\":\"work-key\",\"baseUrl\":\"http://127.0.0.1:27123\"},{\"id\":\"personal\",\"name\":\"Personal\",\"apiKey\":\"personal-key\",\"baseUrl\":\"http://127.0.0.1:27122\"}]"
}Note: For Claude.ai Remote MCP integration, skip this section and use the Tailscale Remote Access Setup instead.
For local MCP clients (e.g., Cline), add to your settings (e.g., cline_mcp_settings.json):
Single-vault configuration:
{
"mcpServers": {
"obsidian-mcp-server": {
"command": "node",
"args": ["/path/to/your/obsidian-mcp-server-enhanced/dist/index.js"],
"env": {
"MCP_AUTH_KEY": "your-generated-mcp-auth-key",
"OBSIDIAN_API_KEY": "your-obsidian-api-key",
"OBSIDIAN_BASE_URL": "http://127.0.0.1:27123",
"OBSIDIAN_ENABLE_CACHE": "true"
}
}
}
}Multi-vault configuration:
{
"mcpServers": {
"obsidian-mcp-server": {
"command": "node",
"args": ["/path/to/your/obsidian-mcp-server-enhanced/dist/index.js"],
"env": {
"MCP_AUTH_KEY": "your-generated-mcp-auth-key",
"OBSIDIAN_VAULTS": "[{\"id\":\"work\",\"name\":\"Work Vault\",\"apiKey\":\"work-api-key\",\"baseUrl\":\"http://127.0.0.1:27123\"},{\"id\":\"personal\",\"name\":\"Personal Vault\",\"apiKey\":\"personal-api-key\",\"baseUrl\":\"http://127.0.0.1:27122\"}]",
"OBSIDIAN_ENABLE_CACHE": "true"
}
}
}
}For remote access to your Obsidian vault from anywhere, you can use Tailscale to securely expose your MCP server over the internet.
- Tailscale Account: Sign up at tailscale.com
- Tailscale Installed: Install Tailscale on your machine running the MCP server
- Tailscale Funnel Enabled: Enable Tailscale Funnel for your account
-
Generate MCP Auth Key:
openssl rand -hex 32
-
Configure Environment: Set up your
.envfile with the generated key:MCP_AUTH_KEY=your-generated-auth-key MCP_TRANSPORT_TYPE=http MCP_HTTP_PORT=3010 MCP_HTTP_STATELESS=true
-
Configure Vaults: Set up single or multi-vault configuration (see Multi-Vault Configuration)
-
Start the MCP Server:
npm run build && npm run start:http -
Enable Tailscale Funnel:
tailscale funnel 3010
-
Get Your Public URL: Check your Tailscale device URL:
tailscale status --self | grep "Funnel on"
Add to your Claude.ai Remote MCP servers using your MCP authentication key:
Single-vault setup:
{
"url": "https://your-device.your-tailnet.ts.net/mcp?api_key=your-mcp-auth-key",
"name": "Obsidian Vault"
}Multi-vault setup:
{
"url": "https://your-device.your-tailnet.ts.net/mcp?api_key=your-mcp-auth-key",
"name": "Obsidian Multi-Vault"
}Authentication Note: Claude.ai Remote MCP uses the
MCP_AUTH_KEYfor server authentication. Individual vault operations use the vault-specific API keys configured in your environment.
- Dual Authentication: MCP server authentication via MCP_AUTH_KEY, vault authentication via individual API keys
- Tailscale Encryption: All traffic is encrypted end-to-end by Tailscale
- Private Network: Only you can access the server through your Tailscale network
- Automatic SSL: Tailscale Funnel provides automatic HTTPS certificates
- Vault Isolation: Each vault uses its own API key for secure access control
Once set up, you can use tools remotely from Claude.ai:
Single-vault commands (uses default vault):
Use obsidian_task_query to show me tasks due today with format="table"
Multi-vault commands (specify vault):
Use obsidian_task_query with vault="work" to show me work tasks due today
Use obsidian_read_file with filePath="daily-note.md" and vault="personal"
Use obsidian_dataview_query with vault="work" to run: TABLE file.name FROM #meeting WHERE file.cday = date(today)
π Pro Tip: For production use, set up automatic startup on boot so your server and Tailscale Funnel start automatically without manual intervention.
Claude connects directly to /mcp, but ChatGPTβs Actions workflow expects an HTTP manifest plus REST-style JSON endpoints. The ChatGPT layer wraps the existing MCP server so you can reuse all transports, vault routing, and task logic without duplicating code.
- Add
CHATGPT_LAYER_ENABLED=trueto your environment (other knobs:CHATGPT_MANIFEST_PATH,CHATGPT_ACTIONS_PATH). - Restart the server (
npm run build && npm run start:http). - Fetch the manifest to confirm availability:
curl "http://127.0.0.1:3010/.well-known/obsidian-chatgpt-manifest.json"
The manifest advertises the MCP endpoint plus a single JSON actions route (/chatgpt/actions). Both endpoints reuse the same MCP_AUTH_KEY query parameter, so you can deploy Tailscale/Claude and ChatGPT in parallel without new auth plumbing.
| Action | Description | Backed By |
|---|---|---|
searchNotes |
Global search with regex/date filters, pagination, and cache fallback. | processObsidianGlobalSearch |
fetchPage |
Fetch markdown or JSON for a note. | ObsidianRestApiService.getFileContent |
updatePage |
Append, prepend, or overwrite a note via whole-file writes. | ObsidianRestApiService methods |
taskQuery |
Tasks plugin query engine with summaries and formatted output. | obsidianTaskQueryLogic |
taskCreate |
Rich task insertion (headings, periodic notes, metadata). | obsidianCreateTaskLogic |
taskUpdate |
Update status/metadata or relocate an existing task. | obsidianUpdateTaskLogic |
All actions accept an optional vault field (defaults to the first configured vault). Requests are JSON payloads of the form:
POST /chatgpt/actions?api_key=YOUR_MCP_AUTH_KEY
{
"action": "searchNotes",
"vault": "work",
"parameters": {
"query": "project alpha",
"searchInPath": "Projects/Alpha",
"modified_since": "last week",
"pageSize": 20
}
}Responses wrap the existing tool result objects:
{
"success": true,
"action": "searchNotes",
"vault": "work",
"data": { "results": [...], "totalFilesFound": 5, "currentPage": 1, ... }
}Use the manifest output inside ChatGPTβs Actions editor (or any HTTP client) to describe the available capabilities while keeping Claude/Tailscale flows unchanged.
The codebase follows a modular structure within the src/ directory:
src/
βββ index.ts # Entry point: Initializes and starts the server
βββ config/ # Configuration loading (env vars, package info)
β βββ index.ts
βββ mcp-server/ # Core MCP server logic and capability registration
β βββ server.ts # Server setup, transport handling, tool/resource registration
β βββ resources/ # MCP Resource implementations (currently none)
β βββ tools/ # MCP Tool implementations (subdirs per tool)
β βββ transports/ # Stdio and HTTP transport logic, auth middleware
βββ services/ # Abstractions for external APIs or internal caching
β βββ obsidianRestAPI/ # Typed client for Obsidian Local REST API
β βββ vaultManager/ # Multi-vault configuration and service management
βββ types-global/ # Shared TypeScript type definitions (errors, etc.)
βββ utils/ # Common utility functions (logger, error handler, security, etc.)
For a detailed file tree, run npm run tree or see docs/tree.md.
This server includes an intelligent in-memory cache designed to enhance performance and resilience when interacting with your vault.
- Performance: By caching file content and metadata, the server can perform search operations much faster, especially in large vaults. This reduces the number of direct requests to the Obsidian Local REST API, resulting in a snappier experience.
- Resilience: The cache acts as a fallback for the
obsidian_global_searchtool. If the live API search fails or times out, the server seamlessly uses the cache to provide results, ensuring that search functionality remains available even if the Obsidian API is temporarily unresponsive. - Efficiency: The cache is designed to be efficient. It performs an initial build on startup and then periodically refreshes in the background by checking for file modifications, ensuring it stays reasonably up-to-date without constant, heavy API polling.
- Initialization: When enabled, the
VaultCacheServicebuilds an in-memory map of all.mdfiles in your vault, storing their content and modification times. - Periodic Refresh: The cache automatically refreshes at a configurable interval (defaulting to 10 minutes). During a refresh, it only fetches content for files that are new or have been modified since the last check.
- Proactive Updates: After a file is modified through a tool like
obsidian_update_file, the service proactively updates the cache for that specific file, ensuring immediate consistency. - Search Fallback: The
obsidian_global_searchtool first attempts a live API search. If this fails, it automatically falls back to searching the in-memory cache.
The cache is enabled by default but can be configured via environment variables:
OBSIDIAN_ENABLE_CACHE: Set totrue(default) orfalseto enable or disable the cache service.OBSIDIAN_CACHE_REFRESH_INTERVAL_MIN: Defines the interval in minutes for the periodic background refresh. Defaults to10.
The Obsidian MCP Server provides a suite of tools for interacting with your vault(s), callable via the Model Context Protocol.
All tools support an optional vault parameter to specify which vault to operate on:
- Default Behavior: Without
vaultparameter, tools use the first configured vault - Specific Vault: Add
vault: "vault-id"to target a specific vault - Example:
obsidian_read_file(filePath="note.md", vault="work")
| Tool Name | Description | Key Arguments |
|---|---|---|
obsidian_read_file |
Retrieves the content and metadata of a file. | filePath, vault?, format?, includeStat? |
obsidian_update_file |
Modifies a file by appending, prepending, or overwriting. | targetType, content, vault?, targetIdentifier?, wholeFileMode |
obsidian_search_replace |
Performs search-and-replace operations in a note. | targetType, replacements, vault?, useRegex?, replaceAll? |
obsidian_global_search |
Searches the entire vault for content. | query, vault?, searchInPath?, useRegex?, page?, pageSize? |
obsidian_list_files |
Lists files and subdirectories in a folder. | dirPath, vault?, fileExtensionFilter?, nameRegexFilter? |
obsidian_manage_frontmatter |
Gets, sets, or deletes keys in a note's frontmatter. | filePath, operation, key, vault?, value? |
obsidian_manage_tags |
Adds, removes, or lists tags in a note. | filePath, operation, tags, vault? |
obsidian_delete_file |
Permanently deletes a file from the vault. | filePath, vault? |
obsidian_dataview_query |
Execute Dataview DQL queries against your vault. | query, vault?, format? |
obsidian_task_query |
Search and analyze tasks across your vault. | vault?, status?, dateRange?, folder?, priority?, format? |
obsidian_periodic_notes |
Create and manage daily, weekly, monthly, yearly notes. | operation, periodType, vault?, date?, content?, append? |
obsidian_block_reference |
Work with block references and heading operations. | operation, filePath, vault?, heading?, content?, blockId? |
obsidian_graph_analysis |
Analyze note connections and vault relationships. | operation, vault?, filePath?, minConnections?, maxDepth? |
obsidian_template_system |
Create files from templates with variable substitution. | operation, vault?, templatePath?, targetPath?, variables? |
obsidian_smart_linking |
Get intelligent link suggestions and recommendations. | operation, vault?, filePath?, content?, maxSuggestions? |
Note: All tools support comprehensive error handling, multi-vault routing, and return structured JSON responses.
This project is licensed under the Apache License 2.0 - see the LICENSE file for details.
This enhanced version is based on the excellent work by cyanheads in the original obsidian-mcp-server project. All core functionality and architecture credit goes to the original author.
Enhancements in this fork:
- Multi-Vault Support: Simultaneous access to multiple Obsidian vaults with vault-specific routing
- Claude.ai Remote MCP integration and compatibility fixes
- Tailscale Funnel integration for secure remote access
- Enhanced HTTP transport layer with authentication separation (MCP_AUTH_KEY)
- Advanced task querying with Tasks plugin integration
- 5 New Advanced Tools: Periodic notes, block references, graph analysis, template system, and smart linking
- Production-ready configuration for enterprise use