🧠 Obsidian MCP Server

A modular Model Context Protocol (MCP) server for interacting with your Obsidian vault from Claude Desktop or your IDE. Navigate, search, create, and analyze your Obsidian notes using natural language commands.

✨ Features

📚 Navigation & Search

• List notes: Explore all notes organized by folders
• Read notes: Get complete content of any note
• Text search: Find specific content across your entire vault
• Date search: Locate notes by modification date range

✍️ Creation & Editing

• Create notes: New notes with metadata and tags
• Add content: Modify existing notes

📊 Analysis

• Vault statistics: Complete analysis of your knowledge
• Metrics: Words, characters, tags, internal links
• Temporal activity: Track your productivity

🚀 Installation

Prerequisites

• Python 3.11+
• uv (dependency manager)
• An Obsidian vault

Installation Steps

1. Clone the repository:

   git clone <your-repository>
   cd obsidian-mcp-server
   

2. Install dependencies:

   uv sync
   

3. Configure your vault:

   cp .env.example .env
   

   Edit .env and set your vault path:

   OBSIDIAN_VAULT_PATH="/path/to/your/obsidian/vault"
   

4. Run the server:

   # Using the new modular entry point (recommended)
   uv run obsidian-mcp-server
   

5. Run tests (optional):

   uv run pytest
   

Quick Setup Script

For automated setup, use the included setup script:

./setup.sh

This script will:

• ✅ Check if UV is installed
• 📦 Install all dependencies
• 🔧 Create .env file from template
• 🧪 Run verification tests
• 📋 Provide next steps

🔧 Configuration

Environment Variables

The .env file must contain:

# Full path to your Obsidian vault
OBSIDIAN_VAULT_PATH="/Users/username/Documents/MyVault"

Claude Desktop Configuration

To use this server with Claude Desktop, add the following configuration to your Claude config file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.jsonWindows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "obsidian-mcp": {
      "command": [
        "uv",
        "run",
        "--directory",
        "/Users/username/path/to/obsidian-mcp-server",
        "obsidian-mcp-server"
      ]
    }
  }
}

(Similar process for other MCP-compatible IDEs like VSCode, Cursor, etc.)

🛠️ Available Tools

📖 Navigation

• listar_notas(carpeta?, incluir_subcarpetas?): List vault notes
• leer_nota(nombre_archivo): Read note content
• buscar_en_notas(texto, carpeta?, solo_titulos?): Search text in notes
• buscar_notas_por_fecha(fecha_desde, fecha_hasta?): Search by date

✏️ Creation & Editing

• crear_nota(titulo, contenido, carpeta?, etiquetas?): Create new note
• agregar_a_nota(nombre_archivo, contenido, al_final?): Add content to existing note

🤖 Prompt Engineering (NEW!)

• guardar_prompt_refinado(titulo, prompt_original, prompt_refinado, contexto, tags?, carpeta?): Save refined prompts for reuse
• actualizar_prompt_biblioteca(nombre_prompt, nuevas_notas?, calificacion?, casos_uso_adicionales?, variaciones?): Update existing prompts
• listar_biblioteca_prompts(): List all saved prompts in your library

📈 Analysis

• estadisticas_vault(): Complete vault statistics

💡 Usage Examples

Once connected to Claude Desktop, you can use commands like:

Basic Operations

• "Show me all my notes about artificial intelligence"
• "Create a note called 'Project Ideas' in the Work folder"
• "What are my vault statistics?"
• "Find notes modified in the last 7 days"
• "Read my note about meditation"

NEW! Prompt Engineering Features

• "Save this refined prompt to my vault" - After Claude creates a great prompt
• "Show me all my saved prompts" - List your prompt library
• "Update my 'Creative Writing' prompt with new usage notes" - Improve existing prompts

Smart Prompt Management Workflow

1. Work with Claude on refining a prompt until it works perfectly
2. Say: "Save this refined prompt as 'Data Analysis Helper' with context about when to use it"
3. Claude usesguardar_prompt_refinado() to save it to your Obsidian vault
4. Later, access your prompt library anytime for reuse and iteration

🧪 Testing

Run the test suite to verify everything works correctly:

# Run all tests
uv run pytest

# Run with verbose output
uv run pytest -v

# Run specific test class
uv run pytest tests/test_basic.py::TestConfiguration

The test suite uses anyio for asynchronous tests and includes:

• ✅ Module imports and dependencies
• ✅ Environment configuration
• ✅ MCP server initialization
• ✅ Vault content validation
• ✅ Tool registration verification
• ✅ Project structure checks

🗂️ Project Structure

Modular Architecture

obsidian-mcp-server/
├── obsidian_mcp/                 # 📦 Main package (modular structure)
│   ├── __init__.py              # Package exports
│   ├── config.py                # ⚙️ Configuration and environment
│   ├── server.py                # 🚀 Main MCP server
│   ├── tools/                   # 🛠️ MCP tools organized by category
│   │   ├── __init__.py
│   │   ├── navigation.py        # 📚 Navigation (list, read, search)
│   │   ├── creation.py          # ✍️ Note creation and editing
│   │   └── analysis.py          # 📊 Analysis and statistics
│   ├── resources/               # 📋 MCP resources
│   │   ├── __init__.py
│   │   └── vault_info.py        # ℹ️ Vault information
│   ├── prompts/                 # 💭 Specialized prompts
│   │   ├── __init__.py
│   │   └── assistant.py         # 🤖 Assistant prompts
│   └── utils/                   # 🔧 Shared utilities
│       ├── __init__.py
│       ├── vault.py             # 📂 Vault utilities
│       └── logging.py           # 📝 Logging configuration
├── main.py                      # 🎯 Main entry point
├── diagnose.py                  # 🔍 Diagnostic script
├── pyproject.toml              # Project configuration
├── pytest.ini                 # Test configuration
├── setup.sh                   # Automated setup script
├── .env                        # Environment variables (not in git)
├── .env.example               # Configuration template
├── .gitignore                 # Git ignore rules
├── README.md                  # This documentation
├── LICENSE                    # MIT License
├── uv.lock                    # Dependency lock file
└── tests/                     # Test suite
    ├── __init__.py
    ├── conftest.py
    └── test_basic.py

Benefits of Modular Architecture

• Separation of concerns: Each module has a specific responsibility
• Maintainability: Organized code that's easy to locate and modify
• Scalability: Easy to add new tools and extend functionality
• Testing: Isolated tests for better coverage and reliability
• Reusability: Components can be imported and used independently

Usage Examples

# Run the server
uv run main.py

# Import modular components
python -c "from obsidian_mcp import create_server; print('✅ Modular import works')"

# Run diagnostics
uv run diagnose.py

🚀 Future Features & Roadmap

Here are some exciting features planned for future releases:

🤖 AI-Powered Intelligence

• Smart Connection Suggestions: AI-powered analysis to suggest meaningful links between related notes based on content similarity and semantic relationships
• Semantic Search: Search by meaning and context, not just exact word matches - find conceptually related content even when different terminology is used
• Question Answering: Ask specific questions about your vault content and get intelligent answers based on your knowledge base

📝 Prompt Engineering Toolkit

• Refined Prompt Library: Automatically save and organize successful prompts from AI conversations for reuse and refinement
• Prompt Templates: Create template prompts for common tasks and scenarios
• Conversation Context Preservation: Save important AI conversation contexts as structured notes

🔍 Advanced Analysis

• Knowledge Graph Visualization: Understand connections and patterns in your knowledge base
• Content Gap Detection: Identify areas where your knowledge could be expanded
• Learning Path Suggestions: AI-recommended reading and study paths based on your current knowledge

🔗 Enhanced Connectivity

• Auto-linking Intelligence: Suggest and create links between conceptually related notes
• Broken Link Management: Detect and help fix broken internal links
• Cross-Reference Analysis: Deep analysis of how concepts relate across your vault

📊 Smart Organization

• Auto-tagging: Intelligent tag suggestions based on content analysis
• Dynamic Organization: AI-suggested folder structures and note categorization
• Duplicate Content Detection: Find and manage similar or duplicate content

These features will make your Obsidian vault not just a storage system, but an intelligent knowledge companion that grows and learns with you.

Want to contribute to any of these features? Check out our contributing guidelines below!

🤝 Contributing

Contributions are welcome! Please:

1. Fork the project
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

📄 License

This project is licensed under the MIT License. See LICENSE for details.

🔗 Useful Links

• FastMCP Documentation
• Model Context Protocol
• Obsidian

⚠️ Notes

• Make sure the Obsidian vault path is correct
• The server requires read/write permissions in the vault directory
• Changes are immediately reflected in Obsidian