obsidian -tools
Add Obsidian integrations like semantic search and custom Templater prompts to Claude or any MCP client.
claude mcp add --transport stdio jacksteamdev-obsidian-mcp-tools docker run -i jacksteamdev/obsidian-mcp-tools-server \ --env VAULT_PATH="path/to/your/vault" \ --env SERVER_LOG_LEVEL="info" \ --env CLAUDE_DESKTOP_URL="https://claude.desktop/api" \ --env LOCAL_REST_API_KEY="your-local-rest-api-key"
How to use
This MCP server acts as a secure bridge between your Obsidian vault and AI applications (such as Claude Desktop). Once the server is running, an MCP client can communicate with your vault through the server, enabling AI-assisted reading of notes, semantic searches, and template execution without giving the AI direct access to your files. The server enforces user consent and configures secure API access, so sensitive vault data remains under your control. When connected to a compatible MCP client, the server exposes standardized endpoints that allow AI tools to request vault content, run templates with dynamic parameters, and perform context-aware searches.
To use the server, install and enable the Obsidian MCP Tools plugin, which will guide you through downloading and starting the local MCP server. The plugin configures Claude Desktop (or other MCP clients) to talk to the local server via a secure API. Through the client, you can ask the AI to read notes for context, perform semantic searches across your vault, or trigger Obsidian templates with values provided by the AI. Remember that features like Vault Access, Semantic Search, and Template Integration only work when the MCP client is connected and you have granted explicit permission for those actions.
How to install
Prerequisites:
- Obsidian installed (v1.7.7 or higher)
- The Obsidian MCP Tools plugin installed and configured
- Local REST API plugin installed and configured with an API key
- Claude Desktop (or another MCP client) installed and set up to communicate with the local server
Installation steps:
- Install the Obsidian MCP Tools plugin from Obsidian's Community Plugins marketplace.
- Open the plugin settings in Obsidian and choose the option to Install Server. This will download and configure the MCP server locally.
- Ensure the Local REST API and Claude Desktop are running and properly configured to talk to the server (the plugin will provide the necessary API keys and endpoints).
- Start the MCP server via the plugin – the server will be exposed on a local address that the MCP client will use (e.g., http://localhost:port).
- In Claude Desktop (or your MCP client), configure the MCP server URL to point to the local server endpoint provided by the plugin. Save and test the connection.
Optional: If you prefer a manual Docker-based run, you can start the server with:
# Example (Docker)
docker run -i jacksteamdev/obsidian-mcp-tools-server
Notes:
- The server binary is distributed as part of the plugin, and logs are written to the plugin's log location.
- The plugin will automatically wire up Claude Desktop with the server; manual configuration is only needed if you bypass the plugin UI.
Additional notes
Tips and troubleshooting:
- Ensure Claude Desktop and the Local REST API plugin are running and configured with valid API keys before starting the MCP server.
- If the server fails to start, check the plugin logs and the client connection status in Claude Desktop. Common issues include permission errors, incorrect API keys, or mismatched server URLs.
- Security: the server is designed to minimize permissions and encrypt communications. Do not expose the local server to the public internet without proper authentication and network restrictions.
- Environment variables can be adjusted to tailor endpoints, vault paths, and logging verbosity. If you encounter issues, try starting with minimal env settings and gradually add extras.
- This MCP server supports vault access, semantic search, and template execution when connected to a compliant MCP client. Ensure your client is configured to request these capabilities explicitly.
Related MCP Servers
Matryoshka
MCP server for token-efficient large document analysis via the use of REPL state
pluggedin-app
The Crossroads for AI Data Exchanges. A unified, self-hostable web interface for discovering, configuring, and managing Model Context Protocol (MCP) servers—bringing together AI tools, workspaces, prompts, and logs from multiple MCP sources (Claude, Cursor, etc.) under one roof.
Agentic -Skill
Agentic-MCP, Progressive MCP client with three-layer lazy loading. Validates AgentSkills.io pattern for efficient token usage. Use MCP without pre-install & wasting full-loading
mongo
MCP server that provide tools to LLMs such as claude in cursor to interact with MongoDB
vscode-context
MCP Server to Connect with VS Code IDE
mcp-turso
MCP server for interacting with Turso-hosted LibSQL databases