home / mcp / obsidian local rest api mcp server
Provides AI-native, task-oriented tools to interact with Obsidian vaults via a local REST API.
Configuration
View docs{
"mcpServers": {
"j-shelfwood-obsidian-local-rest-api-mcp": {
"command": "npx",
"args": [
"obsidian-local-rest-api-mcp"
],
"env": {
"OBSIDIAN_API_KEY": "YOUR_API_KEY_IF_NEEDED",
"OBSIDIAN_API_URL": "http://obsidian-local-rest-api.test"
}
}
}
}This MCP server exposes AI-native, task-focused tools that interact with Obsidian vaults through a local REST API. It is designed to let you reason about high-level tasks rather than low-level file operations, enabling more natural and efficient workflows with LLM clients running locally.
Connect an MCP client to the local Obsidian REST API MCP server to perform intelligent vault tasks. You’ll use high-level tools to explore vaults, read and write notes, manage files, and perform targeted searches. Typical workflows include listing directories with pagination to avoid context overload, upserting notes with frontmatter, retrieving daily or recent notes, performing scoped searches, and discovering related notes to support research or planning.
Prerequisites you need before installing:
Install and run in development or production modes using the supported commands.
Configuration details you’ll set up include environment variables that point the MCP client at the Obsidian API and optionally provide an API key for authentication.
List directory contents with pagination to prevent context overflow, including path, recursion flag, limit, and offset.
Read the contents of any file within the vault by specifying its path.
Write to a file with modes that support replace, append, or prepend to cover create/update scenarios.
Delete a file or directory at a given path.
Intelligent upsert for notes that creates if missing or updates if existing, including frontmatter.
Retrieve a daily note using common naming patterns, such as today’s note.
Fetch notes recently modified, with a limit to control volume.
Perform a multi-scope search with advanced filtering across content, filenames, and tags within an optional path filter.
Discover conceptual relationships between notes based on tags or links to support serendipitous discovery.