obsidian-mcp-plugin  by aaronsb

AI semantic agency for Obsidian knowledge graphs

Created 8 months ago
251 stars

Top 99.8% on SourcePulse

GitHubView on GitHub
Project Summary

This plugin provides AI assistants with direct, semantic access to Obsidian vaults via the Model Context Protocol (MCP). It targets Obsidian users seeking to leverage AI for advanced knowledge graph navigation, concept discovery, and information synthesis, enabling AI to understand and interact with notes as a connected graph rather than isolated files.

How It Works

The plugin transforms Obsidian vaults into a knowledge graph accessible through the open standard MCP. It grants AI "semantic agency" by enabling graph traversal, concept discovery, contextual awareness, and intelligent synthesis across the entire vault. This approach moves beyond single-document AI interaction, allowing AI to follow links, discover related ideas, and combine information semantically from interconnected notes.

Quick Start & Requirements

  • Prerequisites: An MCP-compatible AI client such as Claude Desktop, Claude Code, or Continue.dev.
  • Installation: Via Obsidian Community Plugins (coming soon) or BRAT (for beta testing).
  • Configuration: Requires adding an mcpServers configuration to your AI client, specifying the local server address (e.g., http://localhost:3001/mcp) and optionally authentication details.
  • Documentation: Detailed documentation for tools, features, security, and configuration is available.

Highlighted Details

  • Core Tools: Offers 8 semantic tool groups for comprehensive vault access: vault (file operations), edit (content modification), view (content display), graph (link navigation), workflow (contextual hints), dataview (query execution), bases (database views), and system (server info, web fetch).
  • Semantic Search: Supports advanced query operators (tag:, path:, content:), regular expressions, phrase matching, and relevance ranking.
  • Graph Intelligence: Enables multi-hop traversal with depth control, backlink/forward-link analysis, path finding between concepts, and tag-based navigation.
  • Content Operations: Includes fuzzy text matching for edits, structure-aware modifications, batch operations (split, combine, move), and template support.
  • Integrations: Facilitates Dataview query execution and interaction with Bases database views.

Maintenance & Community

Support is available via GitHub Issues and Discussions. The project accepts sponsorships through GitHub Sponsors.

Licensing & Compatibility

  • License: MIT License.
  • Compatibility: The MIT license generally permits commercial use and integration into closed-source projects.

Limitations & Caveats

The plugin is not yet available in the official Obsidian Community Plugins registry, requiring beta installation via BRAT. The client-side connection requires npx mcp-remote, implying Node.js and npm/yarn dependencies for the AI client setup. Authentication setup requires careful configuration of API keys.

Health Check
Last Commit

3 days ago

Responsiveness

Inactive

Pull Requests (30d)
12
Issues (30d)
1
Star History
30 stars in the last 30 days

Explore Similar Projects

Feedback? Help us improve.