Metadata-Version: 2.4
Name: food-ai
Version: 0.1.0
Summary: An AI agent that monitors computer activity and autonomously decides what food to order.
Author-email: Sagan Labs <hello@sagan-docs.vercel.app>
Project-URL: Homepage, https://github.com/That-Tech-Geek/activity-food-agent
Project-URL: Bug Tracker, https://github.com/That-Tech-Geek/activity-food-agent/issues
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: Microsoft :: Windows
Requires-Python: >=3.10
Description-Content-Type: text/markdown
Requires-Dist: psutil
Requires-Dist: pyyaml
Requires-Dist: requests
Requires-Dist: beautifulsoup4
Requires-Dist: uiautomation
Requires-Dist: pywin32

# Food Agent Activity Monitoring

An AI agent that monitors computer activity and autonomously (or via confirmation) decides what food to order from Swiggy based on your work state.

## Features
- **Cross-platform Activity Sensing**: Detects focused windows (Windows, macOS, Linux).
- **State Classification**: Uses a rolling window to determine if you are in an `INTENSE` work state or a `FUN` relaxed state.
- **Dynamic Scoring**: Different weights for nutrition, convenience, indulgence, and value based on state.
- **SQLite Logging**: Keeps track of all detected states and ordering decisions.
- **Swiggy Placeholder**: Ready for MCP integration.

## Setup
1. Install the agent:
   ```bash
   pip install food-ai
   ```
2. Configure `config.yaml` with your preferred apps and meals.
3. Run the agent:
   ```bash
   python activity_food_agent.py
   ```

## Development
- `activity_food_agent.py`: Main logic.
- `config.yaml`: Configuration and meal database.
- `agent_log.db`: SQLite database for logs.
