📦 AssetKit

A Python toolkit for packaging, discovering, and loading structured runtime assets.

🚀 Features

• ✅ Structured asset packaging with a clean resources/assets/ convention
• ✅ AssetManager: Pythonic runtime asset access interface
• ✅ Optional assets.py auto-generated mapping with dot-access convenience
• ✅ CLI scaffolding for reusable asset packages and app templates
• ✅ Optional asset injection at creation (--add <files/dirs>)
• ✅ Optional --install after generation
• ✅ Optional --gen-assets-py to include reusable assets.py for import
• ✅ Auto-discovery of installed asset packages via entry_points
• ✅ Fully pip-installable — no source directory needed at runtime
• ✅ Supports plain files, binaries, even GitHub repositories

📦 Installation

pip install assetkit

During development:

pip install -e .

🛠 CLI Usage

Create a new asset package:

assetkit new my_assets

Add asset files or folders at creation time:

assetkit new my_assets --add /path/to/data.csv /path/to/config/

Auto-install the package after creation:

assetkit new my_assets --install

Also generate reusable assets.py mapping file:

assetkit new my_assets --gen-assets-py

Put it all together

assetkit new my_assets --add myfile.txt --gen-assets-py --install

Scaffold an AI/ML application project:

assetkit scaffold mlkit my_app_project

📂 Example Asset Package Structure

my_assets/
├── pyproject.toml
├── setup.cfg
├── MANIFEST.in
└── my_assets/
    ├── __init__.py
    ├── assets.py              <-- optional, auto-generated
    └── resources/
        └── assets/
            ├── config/
            │   └── model.yaml
            ├── data/
            │   └── sample.csv
            └── myfile.txt

⚡ Quick Python Usage Example

Manual access via AssetManager:

from assetkit.asset_manager import AssetManager

assets = AssetManager(package_root="my_assets", resource_dir="resources/assets")
print(assets.list())  # List all available assets
print(assets["config/model.yaml"].text())  # Read file contents

Auto-importable mapping via assets.py (if generated):

from my_assets.assets import assets

print(assets.config_model_yaml.text())
print(assets.data_sample_csv.text())
print(assets.myfile_txt.path())  # Full file path

🔍 Discover All Installed Asset Packages

from assetkit.discovery import discover_asset_managers

packages = discover_asset_managers()
for name, assets in packages.items():
    print(f"{name}: {assets.list()}")

🧪 Testing an Installed Asset Package

After creating and installing:

cd my_assets
pip install .

Then test in Python:

from my_assets.assets import assets
print(assets.config_model_yaml.text())

Or with raw AssetManager if no assets.py:

from assetkit import AssetManager
assets = AssetManager(package_root="my_assets", resource_dir="resources/assets")
print(assets.list())

🐳 Dockerized Example (Optional)

FROM python:3.12-slim

RUN pip install assetkit

WORKDIR /app
RUN assetkit new my_assets --add /dev/null --gen-assets-py

WORKDIR /app/my_assets
RUN pip install .

CMD ["python", "-c", "from my_assets.assets import assets; print(assets.myfile_txt.text())"]

📄 License

MIT — See LICENSE

📬 More Info

• GitHub Repository
• PyPI Project Page

🏁 Roadmap (Coming Soon)

• assetkit bundle and assetkit extract CLI tooling
• Language-agnostic assetkit.yaml asset manifests
• assetkit.ext.yaml, assetkit.ext.pandas helpers
• CLI dispatch of packaged binary tools (assetkit run <tool>)
• GitHub Actions publishing workflows
• Plugin-style extension system for asset loaders