AgenticSocial

AgenticSocial is a Python-based application that leverages CrewAI and LitServe to summarize web content and generate engaging social media posts. It uses a combination of language models and tools to fetch, analyze, and summarize web pages, and then craft concise, shareable messages.

Features

• Webpage Summarization: Extracts key insights from any article or webpage.
• Social Media Content Creation: Generates engaging Telegram messages or tweets from the summarized content.
• Configurable: Easily modify settings like the language model, API keys, and server configurations via config.yaml.

Prerequisites

• Python 3.8 or higher
• Required Python packages (see requirements.txt or install dependencies as described below)
• A valid API key for the Firecrawl tool

Installation

1. Clone the repository:

   git clone https://github.com/your-repo/AgenticSocial.git
   cd AgenticSocial

2. Install dependencies:

   pip install -r requirements.txt

3. Set up the configuration file:

   • Create a config.yaml file in the root directory (if it doesn't already exist).
   • Refer to the load_config function in scripts/src/server.py for the default configuration structure.

4. Ensure the data/ directory exists:

   mkdir -p data

Usage

1. Start the server:

   python scripts/src/server.py

2. The server will start on the host and port specified in the config.yaml file (default: 0.0.0.0:8000).

3. Make a POST request to the /predict endpoint with the following JSON payload:

   {
       "url": "https://example.com/some-article"
   }

4. The server will return a JSON response containing:

   • The summarized content
   • A crafted social media message
   • The file path where the results are saved

Configuration

The config.yaml file allows you to customize the following settings:

• Language Model:

  • model: The model name (e.g., qwen2.5)
  • provider: The provider name (e.g., ollama)
  • base_url: The base URL for the language model API

• API Keys:

  • firecrawl: Your Firecrawl API key

• Server:

  • host: The host address for the server
  • port: The port number for the server

Example

To summarize the webpage at https://aws.amazon.com/what-is/reinforcement-learning-from-human-feedback/, send the following request:

curl -X POST http://localhost:8000/predict -H "Content-Type: application/json" -d '{"url": "https://aws.amazon.com/what-is/reinforcement-learning-from-human-feedback/"}'

The response will include the summary, social media message, and the file path where the results are saved.

File Structure

• scripts/src/server.py: Main server script
• data/: Directory where results are saved
• config.yaml: Configuration file (not included by default; create it manually)

License

This project is licensed under the MIT License. See the LICENSE file for details.

Contributing

Contributions are welcome! Feel free to open issues or submit pull requests.

Acknowledgments

• CrewAI for the agent-based framework
• LitServe for the lightweight server framework

curl http://localhost:8080/queue/status

python3 scripts/src/scheduler/processor.py --now