mcpvault and obsidian-mcp-server

These are competing implementations of MCP servers for Obsidian vault access, each offering overlapping functionality (vault reading/querying) with different architectural approaches—users would typically choose one based on feature set and security model rather than use both together.

mcpvault
62
Established
obsidian-mcp-server
54
Established
Maintenance 25/25
Adoption 10/25
Maturity 10/25
Community 17/25
Maintenance 6/25
Adoption 10/25
Maturity 18/25
Community 20/25
Stars: 946
Forks: 70
Downloads:
Commits (30d): 67
Language: Astro
License:
Stars: 396
Forks: 62
Downloads:
Commits (30d): 0
Language: TypeScript
License: Apache-2.0
No License
No risk flags

About mcpvault

bitbonsai/mcpvault

A lightweight Model Context Protocol (MCP) server for safe Obsidian vault access

Provides read/write access to Obsidian vaults through 14 MCP methods including search with BM25 relevance ranking, frontmatter preservation via gray-matter, and tag management, while filtering system directories. Runs as a Node.js stdio transport server compatible with Claude Desktop, ChatGPT, and any MCP-standard AI client, with token-optimized responses and zero external dependencies.

About obsidian-mcp-server

cyanheads/obsidian-mcp-server

Obsidian Knowledge-Management MCP (Model Context Protocol) server that enables AI agents and development tools to interact with an Obsidian vault. It provides a comprehensive suite of tools for reading, writing, searching, and managing notes, tags, and frontmatter, acting as a bridge to the Obsidian Local REST API plugin.

Implements a modular TypeScript architecture leveraging stdio transport for MCP communication with automatic reconnection, backed by an intelligent in-memory vault cache that provides fallback search functionality if the REST API becomes unavailable. Supports flexible file targeting (by path, active note, or periodic notes) and atomic frontmatter operations, with structured logging, Zod validation, and optional HTTP transport via Hono for alternative deployment scenarios.

Scores updated daily from GitHub, PyPI, and npm data. How scores work