Context-aware neural layer with a dynamic tree structure to store and prioritize persistent clues across training epochs.

• 🌳 Context tree with node persistence tracking
• 🧠 Re-evaluation logic for persistent clues based on contextual drift
• 🎯 Hierarchical attention prioritizing stable/persistent nodes
• 🧪 Unit test suite and GitHub Actions CI
• 🛠️ CLI interface (cli.py) for training, exporting, and visualizing
• 🔁 Auto-version bumping via make release
• 📦 PEP 621 compliant pyproject.toml (no setup.py)
• 🧾 Editable install with verification script verify_env.py
• 📘 Docs with Sphinx (make docs)
• 🧰 Makefile for automation

git clone https://github.com/your-username/context-tree-nn-layer.git
cd context-tree-nn-layer
python -m venv venv
source venv/Scripts/activate    # Git Bash
pip install -U pip setuptools wheel
pip install -e .

make verify     # Check Python, pip, editable install
make install    # Install in editable mode
make test       # Run unit tests
make cli        # Run CLI training/export/plot
make docs       # Build docs locally
make release    # Bump version, tag, push

python cli.py --version
python cli.py --train --export --plot

Generate local HTML docs:

make docs
open docs/_build/html/index.html

Or deploy via GitHub Pages: Settings → Pages → Source: /docs

This repo is Skoda-style: clear, concise, modular, and testable. Built with ❤️ by Code Copilot and [you].