Real direction-accuracy/P&L measurement + BULLISH trigger fixes

AI range: containment mode + Target-below-range fix + ML un-floor

Tight directional bands for INTRADAY + 1D

research: feature/selection A/B tools (meta-labeling, sector, delivery, rolling OOS)

Add NVIDIA NIM as free-tier LLM provider; research + UI updates

ML: add indicator + Monte Carlo features, retrain model

ML intraday-high guard, yfinance thread-safety, validation consistency

Fix long-only bias in validation; honest all-prediction hit rate ≥85%

Add dip order probe results CSV; gitignore paper_trading.db.backup (runtime snapshot)

Add dip order probe research script

Fix AI conflicts: persistent retry (no terminal 'AI unavailable'), consistent direction/reasoning, live-anchored 1D/3D targets + gap note

ML model retrain + research updates + news/llm tweaks

Fix empty universe + full-market top picks; drop dead NSE price APIs

Surface all computed indicators (RSI-2/5, EMA20, BB levels, RS-vs-Nifty, Supertrend, EMA-stack, OBV-z, shadow) in AI prompt context

fix: live price AI context, top5 IST midnight reset, remove Primary Call UI

chore: clean up stale docs, add db_backtest script and updated research data

feat: Gemini/SambaNova providers, ATR power-law range model, crash-guard triggers, backtest retry rounds

fix: Ollama timeout 70s, 2-min backoff, safe float in ATR clamp, script fixes

fix: realistic AI ranges, Ollama cold-start warmup, price-target prompt

feat: cost-aware predictions, graded validation, per-TF AI, drop 5D

research: update backtest results and confidence calibration data

refactor: remove GitHub Models + Anthropic; unify LLM provider chain

feat: bearish neutralization guard, T7 trigger, WhatsApp alerts, equity curve + portfolio review

feat: 96% prediction accuracy — regime fix, INTRADAY TF filters, GIFT Nifty, indicators

feat: position averaging, partial close, intraday validation + Nifty regime gate

fix: top5 picker — unblock picks, remove 5-10 day section, enforce 90%+ selection quality

fix: HF provider circuit-breaker, add Groq/HF fallback chain, trigger rules

fix: load dotenv in ai_forecast, improve top5 5D scoring, harden HF repo IDs

Fix prediction validation to use one-sided range check

Add bear-direction fixes, research scripts, market_calendar, and gitignore cleanup