Browser Recall Weekly: Browsing History Analysis and Summarization

Note: This project assumes you're using my Browser Recall project to generate the browsing history database.

Overview

Browser Recall is a Python application designed to analyze your web browsing history, extract meaningful information using a Large Language Model (LLM), generate summaries of your activity, and present the findings in an HTML report. It processes local browser history databases, interacts with an LLM for content analysis, and structures the output for review.

Requirements

• Python 3.x
• Access to an LLM API endpoint compatible with the OpenAI API standard (like LM Studio).
• Browser Recall - This project reads the database generated by my Browser Recall server.

Usage

1. Clone the repository:

   git clone https://github.com/Zetaphor/browser-recall-weekly.git
   cd browser-recall-weekly

2. Install dependencies:

   uv sync

3. Configure src/main.py: Update the configuration constants at the top of the file, especially DB_PATH, API_BASE_URL, MODEL_NAME, and potentially prompt paths. Ensure your LLM server (e.g., LM Studio) is running and the specified model is loaded.
4. Run the application:

   python src/main.py

5. Check the output: Results will be saved in the directory specified by OUTPUT_DIR (default: analysis_results). You should find:
   • A Markdown file with the raw LLM analysis (*_analysis.md).
   • A JSON file with the extracted structured data (*_extracted_data.json).
   • A Markdown file with the browsing summary (*_browsing_summary.md).
   • An HTML report (*_report.html).

Example: