fix(api): add prediction and confidence to /explain response

fix(model): correct preprocessing order and invert heart disease labels

fix(counterfactuals): enable What-If scenarios for heart_disease

refactor(backend): update core inference modules for v2 compatibility

fix: resolve 3 critical integration bugs causing 500 errors

fix: StringDtype bug in engineer_medical_features + retrain model for 16-feature consistency

fix(explain): monkey-patch booster.save_raw() to strip base_score brackets for older SHAP

fix(explain): force SHAP >= 0.50.0 to handle XGBoost 3.x bracket-wrapped base_score

fix: patch XGBoost base_score immediately at model load time (not in explainer property)

fix: correct base_score patch condition in explainer property

fix: add base_score bracket patch in model_loader.py for XGBoost 3.x compatibility

debug: add detailed traceback logging to explain endpoint for diagnosing 500 error

feat(core): apply production security updates, fix infrastructure bugs, and add SHAP service

feat(backend): add bilingual clinical NLP summary for SHAP explanations

feat(backend): resolve model paths relative to project root

feat: implement dynamic disease routing and model loading with Pydantic schemas