Markus, officially known as mcp-obsidian, is an MCP server tool primarily built with Python, enabling AI clients to interact with Obsidian notes. Developed by Markus Pfundstein, this tool integrates with Obsidian via its Local REST API community plugin. It is distributed via pip and also supports one-click installation through McpMux. Categorized under Productivity MCP Servers, Markus offers capabilities for managing Obsidian content, making it a valuable addition for users aiming to automate note-taking and information retrieval tasks using artificial intelligence. It operates over the stdio transport protocol.

What does Markus connect to?

Markus serves as an essential bridge, connecting various AI clients directly to Obsidian. It achieves this integration by utilizing Obsidian's Local REST API community plugin, allowing for programmatic interaction with note vaults. This connectivity enables AI clients like Cursor, Claude Desktop, Claude Code, VS Code, ChatGPT, Windsurf, and JetBrains to access and manipulate Obsidian files. Its design ensures compatibility with any AI client that supports the MCP standard, providing broad utility for users seeking to extend their note management capabilities.

Core features and tools

Markus exposes a set of specific callable tools designed for detailed Obsidian note management. These include list_files_in_vault to view all vault contents, list_files_in_dir for directory-specific listings, and get_file_contents to retrieve note text. Users can also search for documents based on text queries. For content modification, patch_content and append_content allow precise updates or additions to notes. Finally, delete_file provides a way to remove files programmatically, offering a full suite of control over Obsidian data.

Who is it for?

Markus is intended for users who aim to enhance their Obsidian note management through AI. Its capabilities are particularly beneficial for those seeking to automate tasks such as content summarization, information retrieval across their vault, and structured note creation. The tool supports various AI clients, making it suitable for professionals and enthusiasts already using or planning to integrate AI into their personal knowledge management systems. This is not the right fit if you do not use Obsidian for note-taking.

How to get started

The server authenticates using an API key passed as an environment variable. Configure the OBSIDIAN_API_KEY environment variable, and the transport protocol used is stdio.

Use cases

Markus enables several concrete workflows for AI-assisted Obsidian note management, supporting users in automating common tasks. These capabilities extend from simple content retrieval to complex document generation, making daily note interactions more efficient. Specific applications include:

• Summarizing the contents of a specific note to quickly extract key information or generate concise overviews for sharing.
• Searching for documents matching a text query across the entire vault, allowing for rapid information discovery without manual browsing.
• Summarizing meeting notes to create a new note, complete with an introduction tailored for subsequent email distribution.

Markus provides a direct bridge for AI clients to interact with Obsidian notes, offering a suite of tools for content management and retrieval. Its Python foundation and stdio transport make it compatible with numerous AI clients, simplifying AI-driven note automation. The requirement for an Obsidian REST API Key ensures secure access.