Auto-Briefing-Agent

Auto-Briefing-Agent is a Dockerized FastAPI service that scrapes headlines from Hacker News, deduplicates them, and prepares them for a newsletter/LLM-powered briefing pipeline.

It supports scheduled scraping and sending daily briefings via email using LLMs (e.g., Gemini) and Gmail credentials.

Features

• Scrapes top stories from Hacker News
• Deduplicates articles using a SQLite database
• Prepares content for LLM-based summarization
• Sends briefings via email using Gmail SMTP
• Fully containerized with Docker

Project Structure

.
├── app/
│   ├── main.py         # FastAPI application
│   ├── scraper.py      # Hacker News scrapers
│   ├── models.py       # DB models
│   └── database.py     # DB setup
├── tests/              # API & logic tests
├── Dockerfile
├── docker-compose.yml
├── requirements.txt
└── README.md

Prerequisites

Before running the application, ensure you have the following:

1. A Google Gemini API key
2. Gmail credentials

---

Installation

Clone the Repository

git clone https://github.com/MANDRW/Auto-Briefing-Agent.git
cd Auto-Briefing-Agent

Running with Docker

Build and start the services:

docker compose up --build

Starting n8n

Open n8n in your browser:

http://localhost:5678/workflow/

Importing the Workflow

1. Open the n8n UI
2. Click Import Workflow
3. Import the provided workflow.json file
4. Configure the required credentials:
   • Gemini API credentials
   • Gmail credentials
5. Set the recipient email address
6. Save and activate the workflow

API Endpoints

POST /scrape

Scrapes the top 5 articles from Hacker News and returns only new articles (deduplicated by URL).

Request:

curl -X POST http://localhost:8000/scrape

Response:

[
  {
    "id": 1,
    "title": "Article Title",
    "url": "https://example.com/article",
    "created_at": "2024-01-01T12:00:00",
    "is_processed": false
  }
]

GET /health

Health check endpoint.

Response:

{
  "status": "healthy"
}

GET /

Root endpoint with API information.

Database Schema

The Article table contains:

• id: Primary key (auto-increment)
• title: Article title (indexed)
• url: Article URL (unique, indexed)
• created_at: Timestamp of when the article was scraped
• is_processed: Boolean flag for LLM pipeline processing status