Development workflow

# 1. Create a branch
git checkout -b feature/my-improvement

# 2. Make your changes
#    If touching ML features: follow the Three-File Rule

# 3. Run the compatibility check
python3 -c "
import json
mc = json.load(open('models/model_compat.json'))
ml = json.load(open('ml_config.json'))
assert mc['feature_count'] == len(ml['features'])
print('OK')
"

# 3b. Run config authority drift check
python scripts/config_sync.py --check

# 4. Dry-run to confirm nothing is broken
python trading.py --dry

# 5. If you touched ML features: retrain + OOS backtest
python train_ml_model.py --ensemble --position
python backtest_config.py \
  --balance 500 --no-swap --leverage 500 \
  --spread 16.95 --oos-only --no-chart

# 6. Commit
git add <specific files>
git commit -m "feat: <what changed and why>"

.env                   # secrets
models/*.pkl           # ML model binaries (on R2)
models/*.onnx
models/*_xgb.json
models/*_lgb.txt
models/_snapshot_*/    # optimization snapshots
trade_log.csv          # runtime logs
state.json             # runtime state
logs/                  # runtime logs