Personal Context Server

Server converting Limitless AI lifelogs into indexed markdown, searchable via REST API and web UI

Features

• Fetches Lifelog entries
• Persistent vector database using FAISS
• Multiple retrieval modes:
  • Day: Search across entire days
  • Memory: Search within specific memories
  • Section: Search within sections of memories
  • Line: Search individual lines/quotes
• Recency-weighted search results
• Web UI for settings and queries
• REST API for integration

First Time Installation

1. Clone the repository:

   git clone https://github.com/Maclean-D/context-server.git
   cd context-server

2. Get a Limitless API key (Account > Developers)

3. Enter it in config.json

4. Create and activate a virtual environment:

   # Windows
   python -m venv venv
   .\venv\Scripts\activate
   
   # Linux/Mac
   python3 -m venv venv
   source venv/bin/activate

5. Install dependencies:

   pip install -r requirements.txt

6. Start the server:

   python main.py

7. Continute to Starting The Server

Starting The Server

1. Activate a virtual environment:

   # Windows
   .\venv\Scripts\activate
   
   # Linux/Mac
   source venv/bin/activate

2. Start the server:

   python main.py

3. 6. View server status and change settings at http://localhost:5712

Running with Docker

This section provides instructions for running the application using Docker. This is generally recommended as it handles dependencies and environment setup consistently.

Prerequisites:

• Install Docker: Follow the official instructions for your operating system.
  • Install Docker Engine
  • If you plan to use docker-compose, also Install Docker Compose (Docker Desktop for Windows/Mac usually includes it).

Clone the repository (if you haven't already):

git clone https://github.com/Maclean-D/context-server.git
cd context-server

Option 1: Using Docker Compose (Recommended)

Docker Compose simplifies the management of multi-container applications and is convenient for development and deployment.

1. Configure API Key: Create a file named .env in the root of the project directory (alongside docker-compose.yml) with the following content:

   LIMITLESS_API_KEY=your_actual_api_key_here

   Replace your_actual_api_key_here with your actual Limitless API key. If this variable is not set or is empty, the application will attempt to use the API key specified in config.json.

2. Build and run the application: To start the services:

   docker-compose up

   If you have made changes to the application code or the Dockerfile and want to ensure the docker image is rebuilt, use:

   docker-compose up --build

   To run in detached mode (in the background), add the -d flag:

   docker-compose up -d 
   # or with an explicit rebuild
   docker-compose up --build -d

3. Access the application: Open your browser and go to http://localhost:5712.

4. Stopping the application: If running in the foreground, press Ctrl+C. If running in detached mode:

   docker-compose down

Option 2: Using Docker CLI (docker run)

If you prefer not to use Docker Compose, you can build and run the container using Docker CLI commands directly.

1. Prepare Configuration and Directories:

   • Ensure you have a config.json file in the project root. You can copy the provided config.json example or create your own.
   • Create the ./notes and ./faiss_index directories in your project root if they don't exist:

     mkdir -p notes faiss_index

     If config.json does not exist in your project root when you run the container with a volume mount, Docker might create an empty directory instead of a file. It's best to have config.json present.

2. Build the Docker image: Navigate to the project's root directory (where the Dockerfile is located) and run:

   docker build -t context-server-image .

   (You can choose any name for context-server-image).

3. Run the Docker container:

   docker run -d \
     --name context-server-app \
     -p 5712:5712 \
     -v "./data/notes":/app/notes \
     -v "./data/faiss_index":/app/faiss_index \
     -v "./data/config.json":/app/config.json \
     -e LIMITLESS_API_KEY="your_actual_api_key_here" \
     -e UVICORN_RELOAD="false" \
     --restart unless-stopped \
     context-server-image

   Explanation of flags:

   • -d: Run in detached mode (background).
   • --name context-server-app: Assign a name to the container.
   • -p 5712:5712: Map port 5712 on the host to port 5712 in the container.
   • -v "$./data/notes":/app/notes: Mount the local notes directory to /app/notes in the container.
   • -v "$./data/faiss_index":/app/faiss_index: Mount the local faiss_index directory to /app/faiss_index in the container.
   • -v "$./data/config.json":/app/config.json: Mount the local config.json file to /app/config.json in the container.
   • -e LIMITLESS_API_KEY="your_actual_api_key_here": Set the Limitless API key. Replace with your actual key or omit if you want to rely solely on config.json (though setting it here is recommended for docker run).
   • -e UVICORN_RELOAD="false": Ensure Uvicorn's auto-reload is disabled.
   • --restart unless-stopped: Configure the container to restart automatically unless manually stopped.
   • context-server-image: The name of the image you built.

   Replace "your_actual_api_key_here" with your actual Limitless API key.

4. Access the application: Open your browser and go to http://localhost:5712.

5. Viewing logs:

   docker logs context-server-app

6. Stopping the container:

   docker stop context-server-app

7. Removing the container:

   docker rm context-server-app

Data Persistence (for both options): The setup ensures that your data is persisted on your host machine in the following locations within your project directory:

• ./notes: Stores the downloaded markdown lifelogs.
• ./faiss_index: Stores the FAISS vector index.
• ./config.json: Stores your application configuration (when mounted).

These folders will be created in your project directory on your host machine if they don't already exist (though for config.json with docker run, it's best if it exists beforehand).

Using the API

Once your server is started visit http://localhost:5712/docs#/ to view documentation.

Star History

Contributors