Obsidian Diary

You set up an MCP server that runs locally to power AI-assisted journaling in Obsidian. It processes your notes on your machine, builds backlinks, extracts to-dos, and links memory entries to themes, all while keeping data on your device. This makes your journaling workflow faster, more organized, and more interconnected without exposing your data online.

How to use

Install and run the server, then connect your MCP client to it. Use the client to register the diary server, start the local process, and send natural-language prompts to generate memory logs, extract actions, and create links. Typical actions include creating a memory log for today, writing in Obsidian, extracting todos from your Brain Dump, and linking today’s memory with day-backed backlinks.

How to install

Prerequisites you need before starting are Python’s uv package manager, Ollama for local model hosting, an MCP client, and an Obsidian vault.

git clone https://github.com/madebygps/obsidian-diary-mcp.git

cd obsidian-diary-mcp

uv sync

chmod +x start-server.sh

Configuration

Create a local environment file and tailor paths for your Obsidian diary and planner. The following steps assume you will edit environment variables to point to your diary location and planner.

cp .env.example .env

Edit `.env` and set:
- DIARY_PATH: path to your Obsidian diary folder
- PLANNER_PATH: path to your planner data
Optional settings include:
- OLLAMA_MODEL: default llama3.1:latest
- OLLAMA_TIMEOUT: 60s
- OLLAMA_TEMPERATURE: 0.7
- OLLAMA_NUM_PREDICT: 1000

MCP client configuration

Add the server to your MCP client configuration so you can start and communicate with it. Use a short, stable name for the server.

Name: diary
Command: /full/path/to/obsidian-diary-mcp/start-server.sh
Args: []

Available tools

Create memory log

Generate a memory log for a chosen date with prompts based on the past 3 calendar days.

Link memory log

Automatically generate backlinks to a given memory log using date-based citations like [[YYYY-MM-DD]].

Extract todos

Identify action items from brain dumps and output them as organized checklist items.

Show themes

Analyze themes across recent entries and surface related topics.

Memory trace

Analyze memory entries to track theme evolution over time.

Sunday synthesis

Provide weekly synthesis prompts and summaries.