Documentation
Learn how to use SISTER effectively with our comprehensive documentation and guides.
Getting Started
After installing SISTER, you can start using it immediately. Required icons are downloaded automatically as needed.
Optional Setup
To pre-download all icons for offline use:
sister-cli --download
If you've downloaded new icons and need to update the hash cache:
sister-cli --build-hash-cache
Important: The --download command must be run before --build-hash-cache
Usage
SISTER can be used both as a command-line tool and as a Python library for integration into your own projects.
Command Line Usage
# Process a single screenshot
sister-cli -s screenshot.jpg
# Process multiple screenshots
sister-cli -s screenshot1.jpg screenshot2.jpg
# Specify custom output prefix
sister-cli -s screenshot.jpg -o results
# Enable GPU acceleration for OCR (if supported)
sister-cli -s screenshot.jpg --gpu
# Disable automatic image resizing
sister-cli -s screenshot.jpg --no-resize
Python Library Usage
from sister_sto.pipeline.pipeline import build_default_pipeline
pipeline = build_default_pipeline()
pipeline.startup()
result = pipeline.run([image])
pipeline.shutdown()
Configuration
SISTER uses a configuration file located at ~/.sister_sto/config/config.yaml.
Configuration Options
# Default paths can be overridden
data_dir: "~/.sister_sto"
log_dir: "log"
icon_dir: "icons"
overlay_dir: "overlays"
output_dir: "./"
# General settings
log_level: "WARNING"
gpu: false
Custom Config
You can specify a custom config file with the --config option:
sister-cli -s screenshot.jpg --config path/to/config.yaml
Troubleshooting
Poor Recognition Results
- Ensure screenshots are high quality and not heavily compressed
- Verify that the UI scale in STO matches expected sizes
- Check that no UI elements are overlapping important information
- Make sure tooltips or pop-ups aren't covering critical areas
Performance Issues
- Enable GPU acceleration with --gpu if you have compatible hardware
- Reduce logging verbosity if not debugging
- Consider pre-downloading icons when processing many screenshots
- Ensure you have sufficient RAM (8GB minimum recommended)
Installation Problems
- Verify Python 3.11 or higher is installed
- Check system requirements (8GB RAM minimum)
- Install required system libraries for your platform
- Ensure pip is up to date before installation
API Reference
SISTER provides a comprehensive Python API for integration into your own projects.
Pipeline
The main processing pipeline class that orchestrates the recognition process.
View Documentation