Ollama AI document summarizer for PaperlessNGX

This is a document summarization application using BunJS and Ollama AI server to generate AI-powered summaries of documents stored in a Paperless service.

This is a research project on how AI can be used to do useful stuff. Feel free to use it, but use it at your own risk.

Features

• Fetches documents from a specified Paperless URL.
• Generates AI-based summaries using the Ollama server with the specified Llama model.
• Posts the generated summaries back to the Paperless service as comments.
• Supports saving the generated summaries locally as text files.

Requirements

• Node.js and BunJS installed.
• Access to a Paperless service API.
• Running instance of Ollama AI server.

Installation

1. Clone the repository:

   git clone <repository-url>
   cd <project-directory>

1. Install dependencies: Make sure you have BunJS installed. Then run:

   bun install

Configuration

Set up the following environment variables:

• PAPERLESS_TOKEN: Your API token for authenticating with the Paperless service.
• PAPERLESS_URL: The base URL for your Paperless service.
• OUTPUT_TXT: Set to 1 if you want to save summaries as text files, otherwise 0.
• OUTPUT_PATH: Directory path where text summaries should be saved (if OUTPUT_TXT is 1).
• Optional: Override default values for the summarization model settings by setting:
  • MODEL_NAME: Specify the model to use (default is llama3.2).
  • CONTEXT_LENGTH: Set the max context length for the summarizer (default is 8096).
  • SUMMARY_MARKER: Define a custom marker for identifying AI-generated summaries.

Usage

To run the application and start processing documents, execute:

bun run src/index.ts

The script will:

1. Fetch documents from Paperless service without an AI-generated summary.
2. Generate summaries using the configured AI model.
3. Post the generated summaries as comments on each document.
4. Optionally save summaries to the local file system.

Resetting summaries

To reset a summary, simply remove all corresponding notes on the document, that contain your SUMMARY_MARKER, wich is 'AI_SUMMARY' by default.

Notes

• Ensure that your environment variables are correctly set up; otherwise, the script will throw an error.
• Ollama AI server should be running and configured with the required models to generate summaries.

License

This project is open-source and available under the MIT License.

This project was created using bun init in bun v1.1.34. Bun is a fast all-in-one JavaScript runtime.