feat: calibrate account risk scores and threat tiers, resolve clumping in alert queue

feat: implement case access control and guest session dynamic isolation

docs: improve enterprise code documentation, formatting, and 10-agent Copilot architecture details

fix: resolve Mermaid rendering syntax errors, add client auto-fix engine, and cleanup imports

fix: implement pure Python NeighborLoader fallback and weights_only=False in eval script

feat: add hybrid GNN/ML models and improved investigation UI

Data upload feature implemented

fix: graph node laundering annotation, alert queue crash, and popover width shift

fix: prevent call stack overflow in graph investigation

feat: laundering-aware subgraph expansion and copilot updates

feat: copilot action stack, alert popover, conversation persistence, and STR/case actions

feat: enable map pinch zoom, breathe target node pulse, full replay state/camera reversion, and loading screens with AML captions

fix: restore case creation and sidebar polish

fix: database connection pool resilience and copilot case attribution

fix: update copilot redirects to SPA hashes and resolve investigator attribution mapping

fix: standardise case priority to CRITICAL to satisfy cases_priority_check constraint

refactor: integrate Copilot history migration, refine UI action flows, optimize target styles, and clean README

refactor: implement network community overview sampling, fix ruff unused import, and untrack ruff cache

ml: retrain XGBoost from scratch + restore original int(proba*99) scoring

fix: calibrate XGBoost risk scores & classify alert queue by tier

Merge origin/chauhan: bring in UI redesign and frontend enhancements

UI: Align alert cards risk badges and colors, make export buttons green, customize status dropdown select, and clean overview CTAs

fix: recompute GNN feature statistics from dataset at runtime, not checkpoint

fix: build SHAP input DataFrame from pre-coerced dict, not chained indexing

fix: FactCheckingAgent crash on list shap_explanation

fix: GNN NaN scores + SHAP tab animation and invisible title

fix: coerce feature columns to float before XGBoost/SHAP, fix alert enrichment skipping score=0

feat: add RAG retrieval pipeline to AML Copilot

fix: sanitize NaN/Inf floats that crash JSON serialization in investigation endpoint

fix: restore ML models, fix deployment paths, sync alert risk scores with XGBoost

fix: RiskAssessmentAgent crash on list shap_explanation

feat(copilot): support array-based API keys from environment variables

fix: replace shap.TreeExplainer with XGBoost native pred_contribs

fix: build SHAP input as numpy float64 array, not pandas DataFrame

fix: remove simulated XGBoost metrics; harden SHAP against string dtypes

feat: download models from HF Hub at startup; improve GNN trainer

fix: recompute GNN feature statistics from dataset at runtime, not checkpoint

feat(copilot): complete copilot integration and resolve CSS styling compliance

fix(str): resolve pdf font crashes and case linkage filtering

fix: fallback to virtual guest session when Supabase anonymous sign-in is disabled

fix: use boolean values for pinned notes to prevent PostgreSQL type mismatch

feat: integrate case management with Supabase authentication and storage

Improve investigation GNN and metric display

Fix STR export flow

Stabilize investigation SHAP handling

Harden investigation numeric coercion

Fix Vercel proxy and backend dataset fallback

feat: decouple frontend and fix investigation 500 error

feat: Implement caching for API responses and SHAP explanations to improve performance and reduce redundant computations.

Fix investigation view rendering, fix STR PDF generation (unicode/limit), and add graph checkpoint for fast server restart