This is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady
What I Built
SlothSort AI is a safety-focused file organization assistant built with Python and Streamlit. It helps users categorize files into groups such as Documents, Images, Videos, and Archives.
Instead of immediately moving or deleting files, SlothSort AI displays suggested categories first, allowing users to review the proposed organization before taking any action.
The project includes a simple web interface, reusable categorization logic, automated tests, and GitHub Actions checks. It also includes an optional integration path for Ollama and open-weight AI models.
⏱️ The Problem and Potential Time Savings
Manually sorting downloads, documents, images, recordings, and archives can become repetitive and time-consuming.
SlothSort AI aims to reduce this effort by suggesting categories automatically. I have not yet measured actual time savings, so I am presenting this as the project's intended benefit rather than a verified result.
Demo
Try SlothSort AI here:
https://slothsort-ai-4ctucnjvgaqubsivd3kore.streamlit.app/
The live demo includes sample filenames and displays suggested categories. The starter version previews suggestions without moving or deleting files.
Code
GitHub repository:
https://github.com/godGamerz812/SlothSort-AI
The repository includes the Streamlit application, categorization logic, tests, setup instructions, and a GitHub Actions workflow.
How I Built It
Python: Core categorization logic.
Streamlit: Interactive web interface.
File extensions: Fallback categorization for common file types.
Pytest: Automated tests for the categorization logic.
GitHub Actions: Automated Python checks.
Ollama / open-weight AI models: An optional integration path that still needs successful end-to-end inference testing.
I tested the deployed interface with sample filenames and confirmed that the expected categories appear. The optional AI inference has not yet been verified.
Why Does Open Innovation Matter?
Open-source development makes useful tools accessible, inspectable, and adaptable. By sharing SlothSort AI publicly, I hope other developers can explore its implementation, contribute improvements, add new file categories, and help develop a more capable privacy-conscious organization assistant.
The project is designed around a review-first workflow: users should understand the proposed organization before files are changed.
My Agent Session
I used AI assistance during development to help plan the project, build and refine the application, and work through implementation and testing tasks. The repository and deployed demo provide the current project artifacts.
I have not included a DevRelay session recording or claimed successful Gemma/Ollama inference because those have not been verified.
The Lazy Payoff
The goal is simple: spend less time deciding where files belong and more time doing meaningful work.
SlothSort AI is an early, working starter project. Future improvements could include better content-aware categorization, configurable rules, additional file types, and a tested local AI workflow.
Prize Categories
Best Use of Gemma
Abhishek Singhania
Top comments (0)