Munazum is a Python-based command-line tool that analyses and organises cluttered directories (e.g. Downloads) into a clean, predictable structure.
It combines deterministic file rules with assistive machine-learning suggestions while keeping the user fully in control through transparent logging and a strict dry-run mode.
Most file organisers either:
- operate blindly, or
- over-promise opaque “AI automation”.
Munazum prioritises:
- Visibility over automation
- Safety over speed
- Explainability over black-box behaviour
Munazum processes a directory in three stages:
-
Scan Recursively inspects files and subdirectories.
-
Decide Applies rule-based classification, optionally enhanced with confidence-scored ML suggestions.
-
Execute (optional) Copies files into a structured
Organized/directory — only when explicitly permitted.
Organized/
├─ documents/
├─ archives/
├─ executables/
├─ videos/
├─ code/
└─ others/
Original files are never deleted or moved. They are copied only.
git clone https://github.com/AbdullahBahamish/munazum.git
cd munazum
pip install -e .Installed in editable mode for development and global CLI access.
python -m munazum run . --dry-runDisplays planned operations without modifying the filesystem.
python -m munazum run . --dry-run --verboseExample output:
INFO: Target folder: C:\Users\USER\Downloads
INFO: Planned operations: 57
INFO: DRY-RUN: Copying report.pdf → Organized/documents/report.pdf
python -m munazum run .Creates the Organized/ directory and performs the planned copy operations.
- Dry-run mode is strict
- No files or folders are created unless execution is confirmed
- All actions are logged
- No destructive operations by design
- Conservative by default
- Transparent by design
- Modular and extensible
- Built for real-world directories
- Python 3.10+
pyproject.tomlpackaging- Editable installation via
pip - CLI module execution (
python -m munazum)
Abdullah Bahamish Computer Science student · AI & systems enthusiast
MIT License