FileAI

A Python package for organizing and renaming files using AI.

Quick Start

git clone https://github.com/amacinho/fileai.git
cd fileai
pip install .
export GEMINI_API_KEY=<ADD YOUR GEMINI API KEY>
# Check for duplicate files in documents folder (using SHA256 hash of the content)
fileai-dedupe --dry-run ~/tmp/documents
# Actually dedupe files
fileai-dedupe ~/tmp/documents

# Categorize files in documents folder
fileai-process  ~/tmp/documents ~/tmp/output gemini
# fileai will start processing ~/tmp/documents for new files and categorize them under ~/tmp/output.

Configuration

FileAI requires a Gemini API key to function. You can provide your API key in one of three ways:

1. Environment Variable

   export GEMINI_API_KEY=your-api-key

2. Configuration File Create a JSON configuration file at ~/.fileai/config.json (Linux/Mac) or %APPDATA%\fileai\config.json (Windows):

   {
     "api_key": "your-api-key",
     "model": "model-name" # defaults to gemini-2.0-flash-exp
   }

3. Direct Initialization

   from fileai.api import GeminiAPI
   
   api = GeminiAPI(api_key="your-api-key")

The API key precedence is:

1. Direct initialization
2. Configuration file
3. Environment variable

Usage

Command Line Interface

# Basic usage
fileai input_directory output_directory gemini

# With API key provided via command line
fileai input_directory output_directory gemini --api-key your-api-key

# With custom model
fileai input_directory output_directory gemini --model gemini-pro-vision

Python API

from fileai import FileOrganizer
from fileai.api import GeminiAPI

# Initialize the API client (will use config file or environment variables if not provided)
api = GeminiAPI(api_key="your-api-key")  # api_key is optional if configured elsewhere

# Initialize the organizer
organizer = FileOrganizer(input_path="input_dir", output_path="output_dir", api=api)

# Process a single file
organizer.process_file("path/to/your/file.pdf")

# Process a directory
organizer.process_directory("path/to/your/directory")

Features

• Automatically analyzes documents using AI
• Organizes files into appropriate categories
• Generates descriptive filenames based on content
• Supports various file types:
  • PDF documents
  • Images (jpg, png, etc.)
  • Text files
  • Office documents (docx, xlsx, etc.)