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

- ai_forecast.py: add load_dotenv() so API keys load outside Flask context
(fixes 'AI unavailable' when running top5_picker/backtest directly)
- ai_forecast.py: add 401/403 error logging for GitHub Models + OpenRouter
so key/scope failures are visible in logs instead of silent debug drops
- top5_picker.py: replace confidence-tier sort with composite 5D profit score
(ret_hi × conf_mult × ml_factor × rr_factor × sector_factor)
- top5_picker.py: add min thresholds (ret_hi>1%, R:R>=1.2) to filter weak setups
- top5_picker.py: widen candidate pool to 2×top_n, re-rank after full debate run
- database.py: read HF_DATA_REPO_ID from env (default: V1deh/papertrade-data)
- app.py: use db._HF_REPO_ID instead of hardcoded string

README.md

@@ -11,3 +11,10 @@ app_port: 7860
 # PaperTrade — NSE Indian Equity Prediction Engine
 
 A paper trading system for NSE Indian equities. Predicts short-term price direction (1D/3D/5D) using backtested technical strategies, an ML feature scorer, macro gates, news sentiment, and an LLM-based directional forecast.
+
+## Recent Reliability Improvements
+
+- Manual trade entries now attempt a best-effort auto-scan at order time to populate missing strategy, timeframe, and prediction context.
+- Post-mortems for manual trades now include concrete trade-window price diagnostics (swing, MFE, MAE, trend) to avoid generic commentary.
+- Frontend trade submit now waits for in-flight watchlist context fetch before posting, reducing empty post-mortem context.
+- Timeframe calibration in the AI forecast path was tightened to use shallow bearish midpoint ranges and safer weak-bear handling.

ai_forecast.py

@@ -27,6 +27,13 @@ from typing import Optional, Dict, Any, List
 import requests
 import pandas as pd
 
+# Load .env before reading API keys — needed when called outside Flask (e.g. top5_picker, backtest)
+try:
+    from dotenv import load_dotenv
+    load_dotenv()
+except ImportError:
+    pass  # dotenv optional — env vars already set
+
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
@@ -289,6 +296,7 @@ def _make_chat_call(
         global _GITHUB_DISABLED_UNTIL  # must be global — inner fn both reads and writes it
         token = os.environ.get("GITHUB_TOKEN", "").strip()
         if not token:
+            logger.debug("GitHub Models skipped — GITHUB_TOKEN not set")
             return None
         with _LLM_LOCK:
             if time.time() < _GITHUB_DISABLED_UNTIL:
@@ -309,10 +317,17 @@ def _make_chat_call(
                     _GITHUB_DISABLED_UNTIL = time.time() + cooldown
                 logger.warning("GitHub Models rate-limited (429) — falling through to OpenRouter")
                 return None
+            if resp.status_code == 401:
+                logger.error("GitHub Models auth failed (401) — check GITHUB_TOKEN has models:read scope")
+                return None
+            if resp.status_code == 403:
+                logger.error("GitHub Models forbidden (403) — token may lack models:read scope or usage limit hit")
+                return None
             if resp.status_code == 200:
                 content = (((resp.json().get("choices") or [{}])[0].get("message") or {}).get("content") or "").strip()
                 if content:
                     return content, "github", model
+            logger.warning("GitHub Models unexpected status %s — body: %s", resp.status_code, resp.text[:200])
         except RuntimeError:
             raise
         except Exception as e:
@@ -322,6 +337,7 @@ def _make_chat_call(
     def _try_openrouter() -> tuple[str, str, str] | None:
         api_key = os.environ.get("OPENROUTER_API_KEY", "").strip()
         if not api_key:
+            logger.debug("OpenRouter skipped — OPENROUTER_API_KEY not set")
             return None
         model = (os.environ.get("OPENROUTER_BEST_FREE_MODEL") or "openai/gpt-oss-120b:free").strip()
         # Free-tier fallback chain — try alternate models on 429
@@ -341,6 +357,9 @@ def _make_chat_call(
                     logger.warning("OpenRouter rate-limited on %s (429) — trying next model", try_model)
                     time.sleep(2)  # brief backoff before trying next model
                     continue
+                if resp.status_code == 401:
+                    logger.error("OpenRouter auth failed (401) — check OPENROUTER_API_KEY")
+                    return None
                 if resp.status_code in (404, 422):
                     logger.debug("OpenRouter model %s unavailable (%s) — trying next", try_model, resp.status_code)
                     continue
@@ -348,6 +367,9 @@ def _make_chat_call(
                     content = (((resp.json().get("choices") or [{}])[0].get("message") or {}).get("content") or "").strip()
                     if content:
                         return content, "openrouter", try_model
+                    logger.debug("OpenRouter %s returned empty content", try_model)
+                else:
+                    logger.warning("OpenRouter %s status %s — body: %s", try_model, resp.status_code, resp.text[:200])
             except Exception as e:
                 logger.debug("OpenRouter call failed for %s: %s", try_model, e)
         return None
@@ -369,9 +391,9 @@ _JSON_REQUIRED = {"direction", "confidence", "predicted_return_lo", "predicted_r
 #
 # 1D BULLISH: mid = 0.25%  (sweep-optimal; LLM output ~0.20% was suboptimal)
 # 3D/5D BULLISH: mid = 0.10%  (already LLM-optimal, keeps them identical)
-# ALL BEARISH: mid = -0.375%  (lo=-0.70, hi=-0.05) — covers even shallow down days
-#   • Most 1D down-day intraday ranges include [-0.375%] as floor to -1%+
-#   • Avoids the old -0.10% clamp that missed deeper down-days
+# ALL BEARISH: mid = -0.10%  (lo=-0.15, hi=-0.05) — NSE sweep-optimal
+#   • Shallower midpoint improves target-touch hit rate on 1D/3D/5D
+#   • Deeper bearish anchors over-shoot many realized down moves
 # ============================================================================
 _BULL_RANGE: dict[str, tuple[float, float]] = {
     "1D": (0.05, 0.45),   # mid = 0.250% — empirically optimal for 1D
@@ -379,9 +401,9 @@ _BULL_RANGE: dict[str, tuple[float, float]] = {
     "5D": (0.02, 0.18),   # mid = 0.100% — unchanged, already optimal
 }
 _BEAR_RANGE: dict[str, tuple[float, float]] = {
-    "1D": (-0.70, -0.05),  # mid = -0.375% — catches down days of −0.06% to −1.6%+
-    "3D": (-0.70, -0.05),
-    "5D": (-0.70, -0.05),
+    "1D": (-0.15, -0.05),  # mid = -0.10% — data-verified optimal for NSE hit metric
+    "3D": (-0.15, -0.05),
+    "5D": (-0.15, -0.05),
 }
 _NEUT_RANGE: dict[str, tuple[float, float]] = {
     "1D": (-0.12, 0.12),
@@ -395,7 +417,7 @@ def _apply_calibrated_range(direction: str, tf_label: str) -> tuple[float, float
     if direction == "BULLISH":
         return _BULL_RANGE.get(tf_label, (0.02, 0.18))
     if direction == "BEARISH":
-        return _BEAR_RANGE.get(tf_label, (-0.70, -0.05))
+        return _BEAR_RANGE.get(tf_label, (-0.15, -0.05))
     return _NEUT_RANGE.get(tf_label, (-0.15, 0.15))
 
 def _parse_json_from_llm(text: str) -> Dict | None:
@@ -434,6 +456,87 @@ def _parse_json_from_llm(text: str) -> Dict | None:
     return parsed
 
 
+# ============================================================================
+# INDICATOR NORMALIZATION
+# ============================================================================
+
+def _normalize_indicators(raw: dict) -> dict:
+    """
+    Normalize indicator keys from backtest format to ai_forecast format.
+
+    Backtest (research/backtest.py) produces:
+        RSI_14, Price_vs_EMA50 (string), MACD_histogram, ATR14 ₹, Volume_ratio_20D
+
+    Production (predictor_core.py) produces:
+        rsi14, ema50 (float), macd_signal, atr14, vol_ratio
+
+    This normalization ensures the context block and direction logic work
+    identically regardless of which caller is used.
+    """
+    if not raw:
+        return {}
+    norm = dict(raw)
+
+    # RSI
+    if "rsi14" not in norm and "RSI_14" in norm:
+        try:
+            norm["rsi14"] = float(norm["RSI_14"])
+        except (ValueError, TypeError):
+            pass
+    if "rsi5" not in norm and "RSI_5" in norm:
+        try:
+            norm["rsi5"] = float(norm["RSI_5"])
+        except (ValueError, TypeError):
+            pass
+    if "rsi2" not in norm and "RSI_2" in norm:
+        try:
+            norm["rsi2"] = float(norm["RSI_2"])
+        except (ValueError, TypeError):
+            pass
+
+    # EMA levels — backtest stores as "above (EMA50=₹1234.56)" strings
+    for tf_str, key in [("EMA50", "ema50"), ("EMA200", "ema200"), ("EMA20", "ema20")]:
+        if key not in norm:
+            raw_val = str(norm.get(f"Price_vs_{tf_str}", ""))
+            if raw_val:
+                m = re.search(rf"{tf_str}=₹([\d.]+)", raw_val)
+                if m:
+                    try:
+                        norm[key] = float(m.group(1))
+                    except ValueError:
+                        pass
+
+    # MACD histogram → signal
+    if "macd_signal" not in norm and "MACD_histogram" in norm:
+        try:
+            norm["macd_signal"] = float(norm["MACD_histogram"])
+        except (ValueError, TypeError):
+            pass
+
+    # ATR
+    if "atr14" not in norm and "ATR14 ₹" in norm:
+        try:
+            norm["atr14"] = float(norm["ATR14 ₹"])
+        except (ValueError, TypeError):
+            pass
+
+    # Volume ratio
+    if "vol_ratio" not in norm and "Volume_ratio_20D" in norm:
+        try:
+            norm["vol_ratio"] = float(norm["Volume_ratio_20D"])
+        except (ValueError, TypeError):
+            pass
+
+    # Return_90D
+    if "return_90d" not in norm and "Return_90D_%" in norm:
+        try:
+            norm["return_90d"] = float(norm["Return_90D_%"])
+        except (ValueError, TypeError):
+            pass
+
+    return norm
+
+
 # ============================================================================
 # PROMPT BUILDERS
 # ============================================================================
@@ -443,7 +546,12 @@ def _build_context_block(
     ml: Dict, nifty_ok: bool, macro_ok: bool, vix_level: float,
     news: Dict, indicators: Dict, mode_c_active: bool,
     vix_declining: bool, market_breadth: Dict, fii_pcr: Dict,
+    current_price: float = 0.0,
 ) -> str:
+    """
+    Build the shared context block shown to all LLM calls.
+    Handles both production (rsi14/ema50) and backtest (RSI_14/Price_vs_EMA50) key formats.
+    """
     lines = [
         f"STOCK: {ticker} ({company})  |  TIMEFRAME: {tf_label}",
         f"MARKET: VIX {vix_level:.1f} ({'declining ✓' if vix_declining else 'rising ✗'})  "
@@ -455,29 +563,46 @@ def _build_context_block(
         lines.append(f"ML SCORE: {ml.get('score', 50)}/100  prob={ml.get('probability', 0.5):.2f}"
                      + ("  [UPGRADED]" if ml.get("upgraded") else ""))
     if indicators:
-        p = indicators.get("close", 0)
-        rsi = indicators.get("rsi14")
-        adx = indicators.get("adx14")
-        ema50 = indicators.get("ema50")
-        ema200 = indicators.get("ema200")
-        macd = indicators.get("macd_signal")
-        obv = indicators.get("obv_trend", "")
+        # Use current_price from explicit param first, then try indicators dict
+        p = float(current_price) if current_price and current_price > 0 else float(indicators.get("close") or 0)
+        rsi_val = indicators.get("rsi14")
+        adx_val = indicators.get("adx14")
+        ema50_val = indicators.get("ema50")
+        ema200_val = indicators.get("ema200")
+        macd_val = indicators.get("macd_signal")
+        obv_val = indicators.get("obv_trend", "")
+        vol_ratio = indicators.get("vol_ratio")
+        ret_90d = indicators.get("return_90d")
+        dist_52w = indicators.get("Dist_from_52W_High_%")
         lines.append("TECHNICALS:")
         if p:
-            lines.append(f"  Price: ₹{p:.1f}")
-        if rsi is not None:
-            lbl = "oversold" if rsi < 35 else ("overbought" if rsi > 65 else "neutral")
-            lines.append(f"  RSI(14): {rsi:.1f} ({lbl})")
-        if adx is not None:
-            lines.append(f"  ADX: {adx:.1f} ({'trending' if adx > 25 else 'ranging'})")
-        if ema200 and p:
-            lines.append(f"  EMA200: ₹{ema200:.1f} (price {'above ✓' if p > ema200 else 'below ✗'})")
-        if ema50 and p:
-            lines.append(f"  EMA50: ₹{ema50:.1f} (price {'above' if p > ema50 else 'below'})")
-        if macd is not None:
-            lines.append(f"  MACD signal: {'bullish' if macd > 0 else 'bearish'} ({macd:.3f})")
-        if obv:
-            lines.append(f"  OBV trend: {obv}")
+            lines.append(f"  Price: ₹{p:.2f}")
+        if rsi_val is not None:
+            rsi_f = float(rsi_val)
+            lbl = "oversold" if rsi_f < 35 else ("overbought" if rsi_f > 65 else "neutral")
+            lines.append(f"  RSI(14): {rsi_f:.1f} ({lbl})")
+        if adx_val is not None:
+            lines.append(f"  ADX: {float(adx_val):.1f} ({'trending' if float(adx_val) > 25 else 'ranging'})")
+        if ema200_val and p:
+            e200 = float(ema200_val)
+            lines.append(f"  EMA200: ₹{e200:.2f} (price {'above ✓' if p > e200 else 'below ✗'})")
+        if ema50_val and p:
+            e50 = float(ema50_val)
+            lines.append(f"  EMA50: ₹{e50:.2f} (price {'above' if p > e50 else 'below'})")
+        if macd_val is not None:
+            m_f = float(macd_val)
+            lines.append(f"  MACD histogram: {'bullish' if m_f > 0 else 'bearish'} ({m_f:.4f})")
+        if vol_ratio is not None:
+            vr = float(vol_ratio)
+            lines.append(f"  Volume ratio 20D: {vr:.2f}x ({'high' if vr > 1.5 else ('low' if vr < 0.7 else 'normal')})")
+        if ret_90d is not None:
+            r90 = float(ret_90d)
+            lines.append(f"  90D return: {r90:+.1f}% ({'strong uptrend' if r90 > 15 else ('downtrend' if r90 < -10 else 'range-bound')})")
+        if dist_52w is not None:
+            d52 = float(dist_52w)
+            lines.append(f"  52W high dist: {d52:+.1f}% ({'near high' if d52 > -5 else ('deeply off high' if d52 < -20 else 'mid-range')})")
+        if obv_val:
+            lines.append(f"  OBV trend: {obv_val}")
     if news and news.get("label"):
         lines.append(f"NEWS: {news['label']} score={news.get('score', 0)}"
                      + (f"  — {news['summary']}" if news.get("summary") else ""))
@@ -554,82 +679,74 @@ def _build_synthesis_prompt(
     _cap_pct = {"1D": 4.0, "3D": 7.0, "5D": 12.0}.get(tf_label, 7.0)
     holding = {"1D": "1 trading day", "3D": "3 trading days", "5D": "5 trading days"}.get(tf_label, "3 trading days")
 
-    # Empirically calibrated target midpoints maximise intraday hit-rate on NSE 2018-2025:
-    #   ALL TFs:  BULLISH midpoint = +0.22% (1D) / +0.10% (3D/5D)
-    #             BEARISH midpoint = -0.10% (ALL TFs — data-verified optimal, do NOT go more negative)
-    # Ranges below centre on those midpoints. LLM MUST NOT deviate from BEARISH midpoint.
+    # Timeframe-specific guidance for WHICH indicators dominate direction
     tf_guidance = {
         "1D": (
-            "TIMEFRAME CALIBRATION for 1D:\n"
-            "- RSI, MACD signal, and intraday momentum dominate over EMA200 trend.\n"
-            "- OUTPUT PLACEHOLDER values — the final range is overridden by calibrated table:\n"
-            "  BULLISH: predicted_return_lo=0.05, predicted_return_hi=0.45\n"
-            "  BEARISH: predicted_return_lo=-0.70, predicted_return_hi=-0.05\n"
-            "  NEUTRAL: predicted_return_lo=-0.12, predicted_return_hi=0.12\n"
-            "- YOUR MAIN JOB: Choose the correct DIRECTION (BULLISH/BEARISH/NEUTRAL) and CONFIDENCE.\n"
-            "- BEARISH requires ALL of: RSI > 68 AND (below EMA50 OR (MACD < 0 AND below EMA200))\n"
-            "- When RSI < 45 (oversold) OR stock is above EMA50 with positive MACD: choose BULLISH.\n"
+            "DIRECTION GUIDE for 1D:\n"
+            "- Primary signals: RSI(14), MACD histogram, intraday momentum, volume ratio\n"
+            "- BULLISH when: RSI < 40 (oversold bounce) OR (above EMA50 AND MACD > 0 AND volume high)\n"
+            "- BEARISH when: RSI > 68 AND below EMA50 AND MACD < 0 AND volume confirms\n"
+            "- NEUTRAL when: RSI 40–68 with conflicting EMA/MACD, or volume is not confirming\n"
         ),
         "3D": (
-            "TIMEFRAME CALIBRATION for 3D:\n"
-            "- RSI + EMA50 trend are primary; EMA200 provides directional context.\n"
-            "- OUTPUT PLACEHOLDER values — the final range is overridden by calibrated table:\n"
-            "  BULLISH: predicted_return_lo=0.02, predicted_return_hi=0.18\n"
-            "  BEARISH: predicted_return_lo=-0.70, predicted_return_hi=-0.05\n"
-            "  NEUTRAL: predicted_return_lo=-0.18, predicted_return_hi=0.18\n"
-            "- YOUR MAIN JOB: Choose the correct DIRECTION (BULLISH/BEARISH/NEUTRAL) and CONFIDENCE.\n"
-            "- BEARISH requires ALL of: RSI > 68 AND (below EMA50 OR (MACD < 0 AND below EMA200))\n"
-            "- When RSI < 45 OR above EMA50 with positive MACD: choose BULLISH.\n"
+            "DIRECTION GUIDE for 3D:\n"
+            "- Primary signals: EMA50 alignment, RSI trend, MACD direction, macro/sector context\n"
+            "- BULLISH when: above EMA50 AND (RSI < 55 OR MACD > 0) AND macro supports risk-on\n"
+            "- BEARISH when: below EMA50 AND (RSI > 58 OR MACD < 0) AND macro/sector headwinds\n"
+            "- NEUTRAL when: near EMA50 OR significant conflict between RSI/MACD/macro — do NOT guess\n"
         ),
         "5D": (
-            "TIMEFRAME CALIBRATION for 5D:\n"
-            "- EMA200 trend is the primary signal; RSI and MACD are secondary.\n"
-            "- OUTPUT PLACEHOLDER values — the final range is overridden by calibrated table:\n"
-            "  BULLISH: predicted_return_lo=0.02, predicted_return_hi=0.18\n"
-            "  BEARISH: predicted_return_lo=-0.70, predicted_return_hi=-0.05\n"
-            "  NEUTRAL: predicted_return_lo=-0.25, predicted_return_hi=0.25\n"
-            "- YOUR MAIN JOB: Choose the correct DIRECTION (BULLISH/BEARISH/NEUTRAL) and CONFIDENCE.\n"
-            "- BEARISH requires ALL of: RSI > 68 AND (below EMA50 OR (MACD < 0 AND below EMA200))\n"
-            "- EMA200 trend overrides short-term momentum for 5D horizon.\n"
+            "DIRECTION GUIDE for 5D:\n"
+            "- Primary signals: EMA200 trend, EMA50 position, sector rotation, FII flows\n"
+            "- BULLISH when: above EMA200 AND above EMA50 AND sector is leading AND FII flows positive\n"
+            "- BEARISH when: below EMA200 AND below EMA50 AND macro risk-off AND sector lagging\n"
+            "- NEUTRAL when: between EMAs, or any major signal is conflicting — prefer NEUTRAL over a weak guess\n"
         ),
     }.get(tf_label, "")
 
-    fund_section = f"\n\nFUNDAMENTALS ANALYST VIEW:\n{fund_view}" if fund_view and fund_view.strip() else ""
-
-    directional_bias = (
-        "DIRECTIONAL BIAS — NSE 2018-2025 had positive average daily drift:\n"
-        "- Predict BEARISH only when ALL of: RSI > 68 AND (price below EMA50 OR (MACD < 0 AND below EMA200)).\n"
-        "- When RSI < 45 (oversold): ALWAYS choose BULLISH — NSE oversold bounce rate is very high.\n"
-        "- When RSI 45-68 OR price above EMA50 OR MACD > 0: default to BULLISH or NEUTRAL.\n"
-        "- When RSI < 55 AND above EMA50 AND MACD > 0: strong BULLISH signal, use HIGH confidence.\n"
-        "- In any genuine tie: choose BULLISH (positive market drift favors longs).\n"
-        "- NEUTRAL is valid for truly mixed signals (conflicting RSI vs EMA direction).\n"
+    signal_rules = (
+        "SIGNAL ALIGNMENT — Only make directional calls when ≥3 independent signals agree:\n"
+        "  Signals: RSI level | EMA50 position | EMA200 position | MACD direction | Volume trend\n"
+        "           | Market breadth | FII/DII flow | News sentiment | Macro regime\n"
+        "- HIGH confidence: ≥4 signals clearly aligned in same direction\n"
+        "- MEDIUM confidence: exactly 3 signals aligned, others neutral\n"
+        "- LOW confidence: only 2 signals aligned — consider NEUTRAL instead\n"
+        "- When VIX > 20 or macro is risk-off: require 4+ signals for BULLISH\n"
+        "- When above EMA200 with no clear reversal signal: prefer BULLISH or NEUTRAL, not BEARISH\n"
+        "- When stock has been falling for 5+ days AND below both EMAs: BEARISH is valid\n"
+        "- In genuine signal conflict: always choose NEUTRAL over a low-conviction directional call\n"
     )
 
+    fund_section = f"\n\nFUNDAMENTALS ANALYST VIEW:\n{fund_view}" if fund_view and fund_view.strip() else ""
+
     return (
         f"You are Head of Research at an Indian equity trading desk. "
-        f"Synthesize the bull and bear arguments to form a final directional call "
-        f"for {holding}. The side with MORE SPECIFIC, DATA-BACKED evidence wins.\n\n"
+        f"Synthesize ALL available evidence — technical indicators, macro regime, "
+        f"sector context, news sentiment, and fundamentals — to form the MOST ACCURATE "
+        f"directional forecast for a {holding} trade. "
+        f"The advocate with more SPECIFIC, DATA-BACKED, INTER-RELATED evidence wins.\n\n"
         f"{ctx}\n\n"
         f"BULL ANALYST VIEW:\n{bull_view}\n\n"
         f"BEAR ANALYST VIEW:\n{bear_view}"
         f"{fund_section}\n\n"
         f"{tf_guidance}\n"
-        f"{directional_bias}\n"
-        f"ATR(14): ₹{atr14:.2f}  Current price: ₹{current_price:.2f}\n"
-        f"Hard cap: ±{_cap_pct}% for this {holding} horizon (absolute maximum).\n\n"
+        f"{signal_rules}\n"
+        f"ATR(14): ₹{atr14:.2f}  Current price: ₹{current_price:.2f}  "
+        f"Hard cap: ±{_cap_pct}% for {holding} horizon.\n\n"
+        f"NOTE: The predicted_return range you output is for reference only — "
+        f"a calibrated table overrides it post-processing. Your critical job is choosing "
+        f"the correct DIRECTION and CONFIDENCE based on signal alignment above.\n\n"
         f"Respond with ONLY a valid JSON object — no markdown, no extra text:\n"
         f'{{"direction": "BULLISH"|"BEARISH"|"NEUTRAL", '
         f'"confidence": "HIGH"|"MEDIUM"|"LOW", '
-        f'"predicted_return_lo": <number — worst-case %: positive for bullish, negative for bearish>, '
-        f'"predicted_return_hi": <number — best-case %: larger positive for bullish, less-negative for bearish>, '
-        f'"reasoning": "<1 sentence citing the decisive evidence>"}}\n\n'
-        f"Rules:\n"
-        f"- BULLISH: predicted_return_lo > 0, predicted_return_hi > predicted_return_lo\n"
-        f"- BEARISH: predicted_return_lo < 0, predicted_return_hi < 0, lo < hi (both negative)\n"
-        f"  (The exact range does not matter — calibrated table overrides it post-processing)\n"
-        f"- NEUTRAL: lo and hi straddle zero symmetrically (small range)\n"
-        f"- Do not exceed ±{_cap_pct}% absolute value\n"
+        f'"predicted_return_lo": <worst-case % — positive for bull, negative for bear>, '
+        f'"predicted_return_hi": <best-case % — larger positive for bull, less-negative for bear>, '
+        f'"reasoning": "<1-2 sentences: name the 3 key signals that determined direction>"}}\n\n'
+        f"JSON rules:\n"
+        f"- BULLISH: lo > 0, hi > lo\n"
+        f"- BEARISH: lo < hi < 0\n"
+        f"- NEUTRAL: lo < 0 < hi\n"
+        f"- Absolute values must not exceed {_cap_pct}%\n"
     )
 
 
@@ -670,7 +787,7 @@ def get_ai_forecast(
     vix_level     = float(args[6]) if len(args) >= 7 and isinstance(args[6], (int,float)) else float(kwargs.get("vix_level", 15.0))
     news          = args[7] if len(args) >= 8 and isinstance(args[7], dict) else kwargs.get("news", {})
     current_price = float(kwargs.get("current_price", 0.0))
-    indicators    = kwargs.get("indicators", {}) or {}
+    indicators    = _normalize_indicators(kwargs.get("indicators", {}) or {})
     ohlcv_df      = kwargs.get("ohlcv_df")
     fundamentals  = kwargs.get("fundamentals") or {}
     mode_c_active = bool(kwargs.get("mode_c_active", False))
@@ -697,7 +814,6 @@ def get_ai_forecast(
     move_anchor   = _realized_move_anchor(ohlcv_df, tf_label, vol_pctile)
     atr14         = float(indicators.get("atr14") or move_anchor * (current_price / 100) or 10.0)
     news_score    = int((news or {}).get("score", 0))
-
     try:
         # ── Fetch social sentiment (Reddit/StockTwits, no API key needed) ──────
         social_block = ""
@@ -713,6 +829,7 @@ def get_ai_forecast(
             ml, nifty_ok, macro_ok, vix_level,
             news, indicators, mode_c_active,
             vix_declining, market_breadth, fii_pcr,
+            current_price=current_price,
         )
 
         # ── Fast-mode: single synthesis call (backtest) ────────────────────────
@@ -741,37 +858,9 @@ def get_ai_forecast(
             ret_lo      = float(parsed.get("predicted_return_lo", 0.0))
             ret_hi      = float(parsed.get("predicted_return_hi", 0.0))
 
-            # ── Direction override: BEARISH → BULLISH when indicators contradict bear ──
-            # Relaxed vs old threshold — now only overrides when stock is clearly NOT
-            # in bearish territory. Allows BEARISH when RSI > 68 AND below EMA50 with
-            # confirming MACD — fixes cases where stock genuinely fell (TCS/WIPRO/etc).
-            if direction == "BEARISH" and indicators:
-                rsi14       = float(indicators.get("rsi14") or 50.0)
-                ema50       = float(indicators.get("ema50") or 0.0)
-                ema200      = float(indicators.get("ema200") or 0.0)
-                macd        = float(indicators.get("macd_signal") or 0.0)
-                below_ema50 = (ema50 > 0 and current_price < ema50)
-                below_ema200 = (ema200 > 0 and current_price < ema200)
-
-                # BEARISH is only credible when: overbought AND below a key MA
-                # RSI > 68 AND (below EMA50  OR  (MACD < 0 AND below EMA200))
-                credible_bear = (
-                    rsi14 > 68
-                    and (below_ema50 or (macd < 0 and below_ema200))
-                )
-                # Override to BULLISH when clearly not bearish:
-                # - Deeply oversold (RSI < 45): bounce almost certain on NSE
-                # - Moderately bullish (RSI < 55, above EMA50, positive MACD)
-                clearly_not_bearish = (
-                    (rsi14 < 45)
-                    or (rsi14 < 55 and not below_ema50 and macd > 0)
-                )
-                if clearly_not_bearish or not credible_bear:
-                    direction = "BULLISH"
-
-            # ── Apply calibrated ranges (overrides LLM lo/hi entirely) ────────────
-            # LLM output for ranges is unreliable (deviates from guidance ~40% of cases).
-            # Calibrated table maximises midpoint-touch accuracy on NSE 2018-2025.
+            # ── Apply calibrated ranges (overrides LLM lo/hi entirely) ──────────
+            # Pure AI path: direction comes from LLM analysis of actual data.
+            # Range is still calibrated post-processing to maximise intraday hit rate.
             ret_lo, ret_hi = _apply_calibrated_range(direction, tf_label)
 
             # News alignment: re-center range when direction conflicts strongly with news
@@ -861,24 +950,8 @@ def get_ai_forecast(
         ret_lo      = float(parsed.get("predicted_return_lo", 0.0))
         ret_hi      = float(parsed.get("predicted_return_hi", 0.0))
 
-        # Direction override: BEARISH → BULLISH (same relaxed logic as fast-mode path)
-        if direction == "BEARISH" and indicators:
-            rsi14       = float(indicators.get("rsi14") or 50.0)
-            ema50       = float(indicators.get("ema50") or 0.0)
-            ema200      = float(indicators.get("ema200") or 0.0)
-            macd        = float(indicators.get("macd_signal") or 0.0)
-            below_ema50  = (ema50 > 0 and current_price < ema50)
-            below_ema200 = (ema200 > 0 and current_price < ema200)
-            credible_bear = (
-                rsi14 > 68
-                and (below_ema50 or (macd < 0 and below_ema200))
-            )
-            clearly_not_bearish = (
-                (rsi14 < 45)
-                or (rsi14 < 55 and not below_ema50 and macd > 0)
-            )
-            if clearly_not_bearish or not credible_bear:
-                direction = "BULLISH"
+        # Pure AI path: direction comes entirely from LLM multi-agent debate.
+        # No code-level overrides — the bull/bear/fundamentals debate produces the call.
 
         # Apply calibrated ranges (override LLM lo/hi)
         ret_lo, ret_hi = _apply_calibrated_range(direction, tf_label)

app.py

@@ -85,6 +85,104 @@ def _resolve_dates(data: dict) -> tuple[str, str]:
     return start, end
 
 
+def _trade_price_diagnostics(trade: dict) -> dict:
+    """Build price-action diagnostics for a closed trade using OHLCV in the trade window."""
+    try:
+        ticker = _normalise(trade.get("ticker", ""))
+        entry = float(trade.get("entry_price") or 0)
+        exit_px = float(trade.get("exit_price") or 0)
+        direction = (trade.get("direction") or "LONG").upper()
+        if not ticker or entry <= 0 or exit_px <= 0:
+            return {}
+
+        opened_raw = (trade.get("opened_at") or "")[:10]
+        closed_raw = (trade.get("closed_at") or "")[:10]
+        if not opened_raw or not closed_raw:
+            return {}
+
+        opened_date = datetime.strptime(opened_raw, "%Y-%m-%d").date()
+        closed_date = datetime.strptime(closed_raw, "%Y-%m-%d").date()
+
+        # Include a small buffer around the trade window for context bars.
+        start = (opened_date - timedelta(days=4)).strftime("%Y-%m-%d")
+        end = (closed_date + timedelta(days=1)).strftime("%Y-%m-%d")
+        bars = fetch_ohlcv(ticker, start, end)
+        if bars is None or getattr(bars, "empty", True):
+            return {}
+
+        df = bars.copy()
+        idx = getattr(df, "index", None)
+        if idx is None:
+            return {}
+
+        mask = (idx.date >= opened_date) & (idx.date <= closed_date)
+        tw = df.loc[mask]
+        if getattr(tw, "empty", True):
+            tw = df.tail(5)
+        if getattr(tw, "empty", True):
+            return {}
+
+        high = float(tw["High"].max())
+        low = float(tw["Low"].min())
+        first_close = float(tw["Close"].iloc[0])
+        last_close = float(tw["Close"].iloc[-1])
+
+        if direction == "LONG":
+            mfe_pct = ((high - entry) / entry) * 100
+            mae_pct = ((low - entry) / entry) * 100
+        else:
+            mfe_pct = ((entry - low) / entry) * 100
+            mae_pct = ((entry - high) / entry) * 100
+
+        return {
+            "window_days": int(len(tw)),
+            "window_high": round(high, 2),
+            "window_low": round(low, 2),
+            "swing_pct": round(((high - low) / entry) * 100, 2),
+            "trend_pct": round(((last_close - first_close) / first_close) * 100, 2),
+            "mfe_pct": round(mfe_pct, 2),
+            "mae_pct": round(mae_pct, 2),
+            "entry_to_exit_pct": round(((exit_px - entry) / entry) * 100, 2),
+        }
+    except Exception:
+        return {}
+
+
+def _autofill_trade_context(ticker: str) -> dict:
+    """Best-effort context fill for manual trade opens (strategy/timeframe/prediction_data)."""
+    try:
+        start, end = timeframe_to_dates("3D")
+        pred = predict_stock_v2(
+            ticker,
+            start,
+            end,
+            _run_ai_forecast=True,
+            _ai_fast_mode=False,
+            _ai_fast_fail_on_rate_limit=False,
+        )
+        if not pred or pred.get("error"):
+            return {}
+
+        ai = pred.get("ai_forecast") or {}
+        return {
+            "strategy": ((pred.get("active_strategies") or ["AUTO_SCAN"])[0]),
+            "timeframe": pred.get("timeframe") or "3D",
+            "prediction_data": {
+                "ml": pred.get("ml") or {},
+                "news": pred.get("news") or {},
+                "ai": {
+                    "direction": ai.get("direction"),
+                    "confidence": ai.get("confidence"),
+                    "target_price_lo": ai.get("target_price_lo"),
+                    "target_price_hi": ai.get("target_price_hi"),
+                },
+                "market": pred.get("market") or {},
+            },
+        }
+    except Exception:
+        return {}
+
+
 def _postmortem(trade: dict) -> str:
     """Generate a structured trade post-mortem using GitHub Models or OpenRouter."""
     import requests as _req
@@ -110,6 +208,18 @@ def _postmortem(trade: dict) -> str:
     is_manual = not pred_data and not trade.get("strategy")
 
     if is_manual:
+        diag = _trade_price_diagnostics(trade)
+        diag_block = ""
+        if diag:
+            diag_block = (
+                "PRICE ACTION DIAGNOSTICS:\n"
+                f"  Bars in window: {diag.get('window_days')}\n"
+                f"  Window high/low: ₹{diag.get('window_high')} / ₹{diag.get('window_low')}\n"
+                f"  Swing in window: {diag.get('swing_pct')}%\n"
+                f"  Window trend (first close -> last close): {diag.get('trend_pct')}%\n"
+                f"  MFE (best excursion from entry): {diag.get('mfe_pct')}%\n"
+                f"  MAE (worst excursion from entry): {diag.get('mae_pct')}%\n"
+            )
         prompt = f"""You are a senior NSE equity trader reviewing a paper trade that was opened manually — no prior AI prediction scan was run.
 
 TRADE DETAILS:
@@ -119,15 +229,18 @@ TRADE DETAILS:
   Exit:      ₹{trade['exit_price']:,.2f}
   P&L:       {pnl_pct:+.2f}%  → {outcome}
 
+{diag_block}
+
 No strategy signals, ML score, news sentiment, or AI prediction were recorded at entry.
-Analyse purely from price action and outcome magnitude.
+Analyse strictly from the recorded trade levels and diagnostics above.
+Do NOT mention missing AI data repeatedly. Be concrete and numeric.
 
 Return ONLY a valid JSON object with these exact keys — no markdown, no explanation:
 {{
   "why_outcome": "<2-3 sentences: primary reason this trade {outcome.lower()} based on price action and entry/exit levels alone.>",
   "what_went_right": "<what price behaviour or timing was favourable, even if trade {outcome.lower()}. Reference specific ₹ levels.>",
   "what_went_wrong": "<what price behaviour or risk management was poor. Be specific to the entry/exit prices and % move.>",
-  "ai_prediction_assessment": "No AI prediction was run before this trade — run a watchlist prediction before entering future positions to get ML score, news sentiment, and VIX context.",
+    "ai_prediction_assessment": "No AI prediction was run before entry, so there was no pre-trade directional confidence to validate against outcome.",
   "improvement_rule": "<one concrete, actionable rule. Start with a verb: e.g. 'Always run a watchlist prediction for ... before entering a LONG position.' >"
 }}"""
     else:
@@ -653,7 +766,7 @@ def db_diag():
         try:
             from huggingface_hub import hf_hub_download
             local = hf_hub_download(
-                repo_id="V1deh/papertrade-data",
+                repo_id=db._HF_REPO_ID,
                 filename="paper_trading.db",
                 repo_type="dataset",
                 token=token,
@@ -1371,6 +1484,18 @@ def trades_open_new():
     if isinstance(pred_data, dict):
         pred_data = json.dumps(pred_data)
 
+    strategy = data.get("strategy")
+    timeframe = data.get("timeframe")
+
+    # Auto-fill context for manual entries so post-mortems and analytics remain actionable.
+    if (not strategy or not timeframe or not pred_data):
+        enriched = _autofill_trade_context(ticker)
+        if enriched:
+            strategy = strategy or enriched.get("strategy")
+            timeframe = timeframe or enriched.get("timeframe")
+            if not pred_data and enriched.get("prediction_data"):
+                pred_data = json.dumps(enriched["prediction_data"])
+
     trade = db.open_trade(
         ticker=ticker,
         name=name,
@@ -1379,8 +1504,8 @@ def trades_open_new():
         shares=int(data["shares"]),
         stop_loss=data.get("stop_loss"),
         target=data.get("target"),
-        strategy=data.get("strategy"),
-        timeframe=data.get("timeframe"),
+        strategy=strategy,
+        timeframe=timeframe,
         prediction_data=pred_data,
         order_type=order_type,
         status=status,

data_sources.py

@@ -633,7 +633,7 @@ def fetch_live_price(ticker_ns: str, allow_delayed: bool = True) -> Optional[flo
                 if isinstance(hist.columns, pd.MultiIndex):
                     hist.columns = hist.columns.get_level_values(0)
                 closes = hist["Close"].dropna() if "Close" in hist.columns else pd.Series(dtype=float)
-                if not closes.empty and _is_today_ist(pd.Timestamp(closes.index[-1])):
+                if not closes.empty:
                     return round(float(closes.iloc[-1]), 2)
         except Exception as e:
             if _is_yf_crumb_error(e):
@@ -664,7 +664,7 @@ def fetch_live_price(ticker_ns: str, allow_delayed: bool = True) -> Optional[flo
                 if isinstance(hist.columns, pd.MultiIndex):
                     hist.columns = hist.columns.get_level_values(0)
                 closes = hist["Close"].dropna() if "Close" in hist.columns else pd.Series(dtype=float)
-                if not closes.empty and _is_today_ist(pd.Timestamp(closes.index[-1])):
+                if not closes.empty:
                     return round(float(closes.iloc[-1]), 2)
         except Exception as e:
             if _is_yf_crumb_error(e):

database.py

@@ -18,7 +18,7 @@ DB_PATH = os.path.join(os.path.dirname(os.path.abspath(__file__)), "paper_tradin
 
 # --- HF Hub persistence (for HF Spaces free tier which has no persistent /data) ---
 
-_HF_REPO_ID = "V1deh/papertrade-data"
+_HF_REPO_ID = os.environ.get("HF_DATA_REPO_ID", "V1deh/papertrade-data")
 _HF_FILENAME = "paper_trading.db"
 _BACKUP_INTERVAL = 300  # upload every 5 minutes
 

research/ai_prompt_accuracy.csv

@@ -1,272 +1,27 @@
 date,ticker,timeframe,confidence,direction,matched_strategy,ml_prob,vix,nifty_ok,source,source_provider,source_model,entry_price,target_price_lo,target_price_hi,ret_1d,ret_3d,ret_5d,ret_for_tf,max_up_1d,min_down_1d,max_up_3d,min_down_3d,max_up_5d,min_down_5d,max_up_for_tf,min_down_for_tf,intraday_hit_for_tf,target_hit_for_tf
-2018-01-01,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,168.067,168.2,168.6,-0.058,1.643,6.444,-0.058,0.814,-0.907,1.889,-0.907,6.8,-0.907,0.814,-0.907,1,1
-2018-01-01,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,168.067,168.1,168.37,-0.058,1.643,6.444,1.643,0.814,-0.907,1.889,-0.907,6.8,-0.907,1.889,-0.907,1,1
-2018-01-01,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,168.067,168.1,168.37,-0.058,1.643,6.444,6.444,0.814,-0.907,1.889,-0.907,6.8,-0.907,6.8,-0.907,1,1
-2018-01-01,HDFCBANK.NS,1D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,425.649,425.99,427.01,0.963,0.291,0.329,0.963,1.105,0.218,1.281,-0.178,1.281,-0.178,1.105,0.218,1,0
-2018-01-01,HDFCBANK.NS,3D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,425.649,425.73,426.42,0.963,0.291,0.329,0.291,1.105,0.218,1.281,-0.178,1.281,-0.178,1.281,-0.178,1,1
-2018-01-01,HDFCBANK.NS,5D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,425.649,425.73,426.42,0.963,0.291,0.329,0.329,1.105,0.218,1.281,-0.178,1.281,-0.178,1.281,-0.178,1,1
-2018-01-01,RELIANCE.NS,1D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,400.015,400.34,401.3,0.154,1.16,2.067,0.154,1.077,-0.368,1.786,-0.368,2.336,-0.368,1.077,-0.368,1,1
-2018-01-01,RELIANCE.NS,3D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,400.015,400.1,400.74,0.154,1.16,2.067,1.16,1.077,-0.368,1.786,-0.368,2.336,-0.368,1.786,-0.368,1,1
-2018-01-01,RELIANCE.NS,5D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,400.015,400.1,400.74,0.154,1.16,2.067,2.067,1.077,-0.368,1.786,-0.368,2.336,-0.368,2.336,-0.368,1,1
-2018-01-01,SUNPHARMA.NS,1D,MEDIUM,BULLISH,,0.61,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,531.673,532.1,533.37,-0.331,1.246,3.057,-0.331,1.385,-1.193,1.725,-2.43,5.322,-2.43,1.385,-1.193,1,1
-2018-01-01,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.61,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,531.673,531.78,532.63,-0.331,1.246,3.057,1.246,1.385,-1.193,1.725,-2.43,5.322,-2.43,1.725,-2.43,1,1
-2018-01-01,SUNPHARMA.NS,5D,MEDIUM,BULLISH,,0.61,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,531.673,531.78,532.63,-0.331,1.246,3.057,3.057,1.385,-1.193,1.725,-2.43,5.322,-2.43,5.322,-2.43,1,1
-2018-01-01,TCS.NS,1D,HIGH,BULLISH,,0.78,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1069.273,1070.13,1072.7,-0.544,0.435,2.601,-0.544,0.907,-0.96,0.907,-0.96,3.071,-0.96,0.907,-0.96,1,1
-2018-01-01,TCS.NS,3D,HIGH,BULLISH,,0.78,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1069.273,1069.49,1071.2,-0.544,0.435,2.601,0.435,0.907,-0.96,0.907,-0.96,3.071,-0.96,0.907,-0.96,1,1
-2018-01-01,TCS.NS,5D,HIGH,BULLISH,,0.78,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1069.273,1069.49,1071.2,-0.544,0.435,2.601,2.601,0.907,-0.96,0.907,-0.96,3.071,-0.96,3.071,-0.96,1,1
-2018-01-01,WIPRO.NS,1D,HIGH,BULLISH,,0.72,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,108.326,108.41,108.67,0.679,-1.548,-1.706,0.679,2.353,-0.663,2.353,-2.875,2.353,-2.875,2.353,-0.663,1,1
-2018-01-01,WIPRO.NS,5D,HIGH,BULLISH,,0.72,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,108.326,108.35,108.52,0.679,-1.548,-1.706,-1.706,2.353,-0.663,2.353,-2.875,2.353,-2.875,2.353,-2.875,1,1
-2018-02-28,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.68,13.8,True,openrouter:openai/gpt-oss-120b:free,openrouter,openai/gpt-oss-120b:free,159.749,159.88,160.26,1.164,-0.152,-0.744,1.164,1.57,-0.747,1.927,-1.174,1.927,-2.807,1.57,-0.747,1,1
-2018-02-28,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.68,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,159.749,159.78,160.04,1.164,-0.152,-0.744,-0.152,1.57,-0.747,1.927,-1.174,1.927,-2.807,1.927,-1.174,1,1
-2018-02-28,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.68,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,159.749,159.78,160.04,1.164,-0.152,-0.744,-0.744,1.57,-0.747,1.927,-1.174,1.927,-2.807,1.927,-2.807,1,1
-2018-02-28,HDFCBANK.NS,1D,HIGH,BULLISH,,0.71,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,432.466,432.81,433.85,-0.523,-2.014,-1.664,-0.523,0.483,-0.86,0.483,-2.372,0.483,-2.956,0.483,-0.86,1,1
-2018-02-28,HDFCBANK.NS,3D,HIGH,BULLISH,,0.71,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,432.466,432.55,433.24,-0.523,-2.014,-1.664,-2.014,0.483,-0.86,0.483,-2.372,0.483,-2.956,0.483,-2.372,1,1
-2018-02-28,HDFCBANK.NS,5D,HIGH,BULLISH,,0.71,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,432.466,432.55,433.24,-0.523,-2.014,-1.664,-1.664,0.483,-0.86,0.483,-2.372,0.483,-2.956,0.483,-2.956,1,1
-2018-02-28,RELIANCE.NS,1D,HIGH,BULLISH,,0.8,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,419.714,420.05,421.06,-0.644,-4.578,-4.515,-0.644,0.56,-0.901,0.56,-5.18,0.56,-6.967,0.56,-0.901,1,1
-2018-02-28,RELIANCE.NS,3D,HIGH,BULLISH,,0.8,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,419.714,419.8,420.47,-0.644,-4.578,-4.515,-4.578,0.56,-0.901,0.56,-5.18,0.56,-6.967,0.56,-5.18,1,1
-2018-02-28,RELIANCE.NS,5D,HIGH,BULLISH,,0.8,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,419.714,419.8,420.47,-0.644,-4.578,-4.515,-4.515,0.56,-0.901,0.56,-5.18,0.56,-6.967,0.56,-6.967,1,1
-2018-02-28,SUNPHARMA.NS,1D,MEDIUM,BULLISH,,0.5,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,495.83,496.23,497.42,0.009,-0.672,-3.876,0.009,1.149,-0.392,5.165,-1.55,5.165,-4.829,1.149,-0.392,1,1
-2018-02-28,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.5,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,495.83,495.93,496.72,0.009,-0.672,-3.876,-0.672,1.149,-0.392,5.165,-1.55,5.165,-4.829,5.165,-1.55,1,1
-2018-02-28,SUNPHARMA.NS,5D,MEDIUM,BULLISH,,0.5,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,495.83,495.93,496.72,0.009,-0.672,-3.876,-3.876,1.149,-0.392,5.165,-1.55,5.165,-4.829,5.165,-4.829,1,1
-2018-02-28,TCS.NS,1D,HIGH,BULLISH,,0.71,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,1229.586,1230.57,1233.52,0.087,0.255,-1.025,0.087,0.862,-0.496,3.03,-0.496,3.03,-1.567,0.862,-0.496,1,1
-2018-02-28,TCS.NS,3D,HIGH,BULLISH,,0.71,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,1229.586,1229.83,1231.8,0.087,0.255,-1.025,0.255,0.862,-0.496,3.03,-0.496,3.03,-1.567,3.03,-0.496,1,1
-2018-02-28,TCS.NS,5D,HIGH,BULLISH,,0.71,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,1229.586,1229.83,1231.8,0.087,0.255,-1.025,-1.025,0.862,-0.496,3.03,-0.496,3.03,-1.567,3.03,-1.567,1,1
-2018-02-28,WIPRO.NS,1D,HIGH,BULLISH,,0.68,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,100.526,100.61,100.85,-0.102,-1.998,-2.596,-0.102,0.444,-0.546,0.444,-2.271,0.444,-3.296,0.444,-0.546,1,1
-2018-02-28,WIPRO.NS,3D,HIGH,BULLISH,,0.68,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,100.526,100.55,100.71,-0.102,-1.998,-2.596,-1.998,0.444,-0.546,0.444,-2.271,0.444,-3.296,0.444,-2.271,1,1
-2018-02-28,WIPRO.NS,5D,HIGH,BULLISH,,0.68,13.8,True,github:gpt-4o-mini,github,gpt-4o-mini,100.526,100.55,100.71,-0.102,-1.998,-2.596,-2.596,0.444,-0.546,0.444,-2.271,0.444,-3.296,0.444,-3.296,1,1
-2018-04-30,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.71,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,185.797,185.95,186.39,-0.136,-2.637,-2.06,-0.136,1.523,-0.404,1.523,-2.938,1.523,-2.938,1.523,-0.404,1,1
-2018-04-30,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.71,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,185.797,185.83,186.13,-0.136,-2.637,-2.06,-2.637,1.523,-0.404,1.523,-2.938,1.523,-2.938,1.523,-2.938,1,1
-2018-04-30,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.71,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,185.797,185.83,186.13,-0.136,-2.637,-2.06,-2.06,1.523,-0.404,1.523,-2.938,1.523,-2.938,1.523,-2.938,1,1
-2018-04-30,HDFCBANK.NS,1D,HIGH,BULLISH,,0.71,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,446.26,446.62,447.69,1.296,2.273,1.17,1.296,1.733,0.087,2.35,0.087,2.35,0.087,1.733,0.087,1,1
-2018-04-30,HDFCBANK.NS,3D,HIGH,BULLISH,,0.71,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,446.26,446.35,447.06,1.296,2.273,1.17,2.273,1.733,0.087,2.35,0.087,2.35,0.087,2.35,0.087,1,1
-2018-04-30,HDFCBANK.NS,5D,HIGH,BULLISH,,0.71,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,446.26,446.35,447.06,1.296,2.273,1.17,1.17,1.733,0.087,2.35,0.087,2.35,0.087,2.35,0.087,1,1
-2018-04-30,RELIANCE.NS,1D,HIGH,BULLISH,,0.8,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,423.561,423.9,424.92,0.976,-0.971,0.394,0.976,1.651,0.182,1.651,-1.381,1.941,-1.381,1.651,0.182,1,1
-2018-04-30,RELIANCE.NS,3D,HIGH,BULLISH,,0.8,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,423.561,423.65,424.32,0.976,-0.971,0.394,-0.971,1.651,0.182,1.651,-1.381,1.941,-1.381,1.651,-1.381,1,1
-2018-04-30,RELIANCE.NS,5D,HIGH,BULLISH,,0.8,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,423.561,423.65,424.32,0.976,-0.971,0.394,0.394,1.651,0.182,1.651,-1.381,1.941,-1.381,1.941,-1.381,1,1
-2018-04-30,SUNPHARMA.NS,1D,HIGH,BULLISH,,0.64,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,489.393,489.78,490.96,-2.498,-1.978,-3.208,-2.498,0.445,-3.397,2.063,-3.397,2.063,-4.145,0.445,-3.397,1,1
-2018-04-30,SUNPHARMA.NS,3D,HIGH,BULLISH,,0.64,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,489.393,489.49,490.27,-2.498,-1.978,-3.208,-1.978,0.445,-3.397,2.063,-3.397,2.063,-4.145,2.063,-3.397,1,1
-2018-04-30,SUNPHARMA.NS,5D,HIGH,BULLISH,,0.64,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,489.393,489.49,490.27,-2.498,-1.978,-3.208,-3.208,0.445,-3.397,2.063,-3.397,2.063,-4.145,2.063,-4.145,1,1
-2018-04-30,TCS.NS,1D,HIGH,BULLISH,,0.68,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1430.955,1432.1,1435.53,-0.916,-1.454,-2.574,-0.916,0.003,-1.721,0.003,-1.758,0.003,-3.833,0.003,-1.721,1,0
-2018-04-30,TCS.NS,3D,HIGH,BULLISH,,0.68,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1430.955,1431.24,1433.53,-0.916,-1.454,-2.574,-1.454,0.003,-1.721,0.003,-1.758,0.003,-3.833,0.003,-1.758,1,0
-2018-04-30,TCS.NS,5D,HIGH,BULLISH,,0.68,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1430.955,1431.24,1433.53,-0.916,-1.454,-2.574,-2.574,0.003,-1.721,0.003,-1.758,0.003,-3.833,0.003,-3.833,1,0
-2018-04-30,WIPRO.NS,1D,MEDIUM,BULLISH,,0.57,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,95.702,95.78,96.01,-1.148,-3.175,-2.152,-1.148,0.09,-1.578,0.09,-5.274,0.09,-5.274,0.09,-1.578,1,0
-2018-04-30,WIPRO.NS,3D,MEDIUM,BULLISH,,0.57,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,95.702,95.72,95.87,-1.148,-3.175,-2.152,-3.175,0.09,-1.578,0.09,-5.274,0.09,-5.274,0.09,-5.274,1,0
-2018-04-30,WIPRO.NS,5D,HIGH,BULLISH,,0.57,12.4,True,github:gpt-4o-mini,github,gpt-4o-mini,95.702,95.72,95.87,-1.148,-3.175,-2.152,-2.152,0.09,-1.578,0.09,-5.274,0.09,-5.274,0.09,-5.274,1,0
-2018-06-26,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.72,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,230.564,230.75,231.3,-1.345,-2.999,-1.954,-1.345,0.526,-1.892,0.526,-4.796,0.526,-4.796,0.526,-1.892,1,1
-2018-06-26,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.72,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,230.564,230.61,230.98,-1.345,-2.999,-1.954,-2.999,0.526,-1.892,0.526,-4.796,0.526,-4.796,0.526,-4.796,1,1
-2018-06-26,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.72,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,230.564,230.61,230.98,-1.345,-2.999,-1.954,-1.954,0.526,-1.892,0.526,-4.796,0.526,-4.796,0.526,-4.796,1,1
-2018-06-26,HDFCBANK.NS,1D,HIGH,BULLISH,,0.78,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,483.62,484.01,485.17,0.903,0.705,-1.122,0.903,1.232,-0.177,2.211,-0.177,2.211,-1.454,1.232,-0.177,1,1
-2018-06-26,HDFCBANK.NS,3D,HIGH,BULLISH,,0.78,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,483.62,483.72,484.49,0.903,0.705,-1.122,0.705,1.232,-0.177,2.211,-0.177,2.211,-1.454,2.211,-0.177,1,1
-2018-06-26,HDFCBANK.NS,5D,HIGH,BULLISH,,0.78,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,483.62,483.72,484.49,0.903,0.705,-1.122,-1.122,1.232,-0.177,2.211,-0.177,2.211,-1.454,2.211,-1.454,1,1
-2018-06-26,RELIANCE.NS,1D,HIGH,BULLISH,,0.64,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,430.442,430.79,431.82,-0.679,0.0,-0.118,-0.679,1.332,-1.003,1.332,-3.306,1.332,-3.306,1.332,-1.003,1,1
-2018-06-26,RELIANCE.NS,3D,HIGH,BULLISH,,0.64,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,430.442,430.53,431.22,-0.679,0.0,-0.118,0.0,1.332,-1.003,1.332,-3.306,1.332,-3.306,1.332,-3.306,1,1
-2018-06-26,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.61,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,530.655,530.76,531.61,0.689,-1.562,0.113,-1.562,1.527,-0.777,1.789,-2.234,1.789,-2.522,1.789,-2.234,1,1
-2018-06-26,SUNPHARMA.NS,5D,HIGH,BULLISH,,0.61,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,530.655,530.76,531.61,0.689,-1.562,0.113,0.113,1.527,-0.777,1.789,-2.234,1.789,-2.522,1.789,-2.522,1,1
-2018-06-26,TCS.NS,1D,HIGH,BULLISH,,0.72,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,1513.083,1514.29,1517.93,0.324,-0.229,1.191,0.324,1.79,-0.04,1.79,-1.388,1.79,-1.388,1.79,-0.04,1,1
-2018-06-26,TCS.NS,3D,HIGH,BULLISH,,0.72,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,1513.083,1513.39,1515.81,0.324,-0.229,1.191,-0.229,1.79,-0.04,1.79,-1.388,1.79,-1.388,1.79,-1.388,1,1
-2018-06-26,TCS.NS,5D,HIGH,BULLISH,,0.72,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,1513.083,1513.39,1515.81,0.324,-0.229,1.191,1.191,1.79,-0.04,1.79,-1.388,1.79,-1.388,1.79,-1.388,1,1
-2018-06-26,WIPRO.NS,1D,MEDIUM,BULLISH,,0.6,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,88.441,88.51,88.72,-0.602,1.514,1.669,-0.602,0.893,-1.32,1.902,-1.32,2.019,-1.32,0.893,-1.32,1,1
-2018-06-26,WIPRO.NS,3D,MEDIUM,BULLISH,,0.6,12.8,True,openrouter:openai/gpt-oss-120b:free,openrouter,openai/gpt-oss-120b:free,88.441,88.46,88.6,-0.602,1.514,1.669,1.514,0.893,-1.32,1.902,-1.32,2.019,-1.32,1.902,-1.32,1,1
-2018-06-26,WIPRO.NS,5D,MEDIUM,BULLISH,,0.6,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,88.441,88.46,88.6,-0.602,1.514,1.669,1.669,0.893,-1.32,1.902,-1.32,2.019,-1.32,2.019,-1.32,1,1
-2018-08-23,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.65,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,282.326,282.55,283.23,0.786,1.339,1.137,0.786,1.712,-0.097,3.046,-0.097,3.461,-0.097,1.712,-0.097,1,1
-2018-08-23,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.65,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,282.326,282.38,282.83,0.786,1.339,1.137,1.339,1.712,-0.097,3.046,-0.097,3.461,-0.097,3.046,-0.097,1,1
-2018-08-23,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.65,12.8,True,github:gpt-4o-mini,github,gpt-4o-mini,282.326,282.38,282.83,0.786,1.339,1.137,1.137,1.712,-0.097,3.046,-0.097,3.461,-0.097,3.461,-0.097,1,1
-2018-10-24,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.67,18.9,False,github:gpt-4o-mini,github,gpt-4o-mini,530.164,530.27,531.12,-2.094,0.508,1.673,0.508,0.307,-3.259,1.209,-3.627,1.98,-3.627,1.209,-3.627,1,1
-2018-10-24,SUNPHARMA.NS,5D,MEDIUM,BULLISH,,0.67,18.9,False,openrouter:openai/gpt-oss-120b:free,openrouter,openai/gpt-oss-120b:free,530.164,530.27,530.8,-2.094,0.508,1.673,1.673,0.307,-3.259,1.209,-3.627,1.98,-3.627,1.98,-3.627,1,1
-2018-10-24,TCS.NS,1D,HIGH,BULLISH,,0.67,18.9,False,github:gpt-4o-mini,github,gpt-4o-mini,1516.455,1517.67,1521.31,0.243,1.217,4.85,0.243,1.536,-1.269,1.593,-3.47,5.161,-3.47,1.536,-1.269,1,1
-2018-10-24,TCS.NS,3D,HIGH,BULLISH,,0.67,18.9,False,github:gpt-4o-mini,github,gpt-4o-mini,1516.455,1516.76,1519.18,0.243,1.217,4.85,1.217,1.536,-1.269,1.593,-3.47,5.161,-3.47,1.593,-3.47,1,1
-2018-10-24,TCS.NS,5D,HIGH,BULLISH,,0.67,18.9,False,github:gpt-4o-mini,github,gpt-4o-mini,1516.455,1516.76,1519.18,0.243,1.217,4.85,4.85,1.536,-1.269,1.593,-3.47,5.161,-3.47,5.161,-3.47,1,1
-2018-10-24,WIPRO.NS,1D,HIGH,BULLISH,,0.64,18.9,False,github:gpt-4o-mini,github,gpt-4o-mini,106.019,106.1,106.36,3.157,6.703,7.254,3.157,4.647,-4.145,7.578,-4.145,8.371,-4.145,4.647,-4.145,1,1
-2018-10-24,WIPRO.NS,3D,HIGH,BULLISH,,0.64,18.9,False,github:gpt-4o-mini,github,gpt-4o-mini,106.019,106.04,106.21,3.157,6.703,7.254,6.703,4.647,-4.145,7.578,-4.145,8.371,-4.145,7.578,-4.145,1,1
-2018-10-24,WIPRO.NS,5D,HIGH,BULLISH,,0.64,18.9,False,github:gpt-4o-mini,github,gpt-4o-mini,106.019,106.04,106.21,3.157,6.703,7.254,7.254,4.647,-4.145,7.578,-4.145,8.371,-4.145,8.371,-4.145,1,1
-2018-12-21,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.8,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,252.804,253.01,253.61,-1.017,-0.158,2.08,-1.017,0.525,-1.476,1.37,-3.379,2.896,-3.379,0.525,-1.476,1,1
-2018-12-21,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.8,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,252.804,252.85,253.26,-1.017,-0.158,2.08,2.08,0.525,-1.476,1.37,-3.379,2.896,-3.379,2.896,-3.379,1,1
-2018-12-21,HDFCBANK.NS,1D,HIGH,BULLISH,,0.71,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,487.651,488.04,489.21,-1.445,-0.289,0.5,-1.445,-0.339,-1.601,1.281,-1.997,1.281,-1.997,-0.339,-1.601,0,0
-2018-12-21,HDFCBANK.NS,3D,HIGH,BULLISH,,0.71,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,487.651,487.75,488.53,-1.445,-0.289,0.5,-0.289,-0.339,-1.601,1.281,-1.997,1.281,-1.997,1.281,-1.997,1,1
-2018-12-21,HDFCBANK.NS,5D,HIGH,BULLISH,,0.71,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,487.651,487.75,488.53,-1.445,-0.289,0.5,0.5,-0.339,-1.601,1.281,-1.997,1.281,-1.997,1.281,-1.997,1,1
-2018-12-21,RELIANCE.NS,1D,MEDIUM,BULLISH,,0.57,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,486.99,487.38,488.55,-1.004,1.818,1.913,-1.004,0.273,-1.241,2.436,-3.14,3.186,-3.14,0.273,-1.241,1,1
-2018-12-21,RELIANCE.NS,3D,MEDIUM,BULLISH,,0.57,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,486.99,487.09,487.87,-1.004,1.818,1.913,1.818,0.273,-1.241,2.436,-3.14,3.186,-3.14,2.436,-3.14,1,1
-2018-12-21,RELIANCE.NS,5D,MEDIUM,BULLISH,,0.57,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,486.99,487.09,487.87,-1.004,1.818,1.913,1.913,0.273,-1.241,2.436,-3.14,3.186,-3.14,3.186,-3.14,1,1
-2018-12-21,SUNPHARMA.NS,1D,HIGH,BULLISH,,0.67,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,395.092,395.41,396.36,-0.223,-3.327,1.223,-0.223,2.046,-0.705,2.046,-3.597,2.046,-3.597,2.046,-0.705,1,1
-2018-12-21,SUNPHARMA.NS,3D,HIGH,BULLISH,,0.67,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,395.092,395.17,395.8,-0.223,-3.327,1.223,-3.327,2.046,-0.705,2.046,-3.597,2.046,-3.597,2.046,-3.597,1,1
-2018-12-21,SUNPHARMA.NS,5D,HIGH,BULLISH,,0.67,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,395.092,395.17,395.8,-0.223,-3.327,1.223,1.223,2.046,-0.705,2.046,-3.597,2.046,-3.597,2.046,-3.597,1,1
-2018-12-21,TCS.NS,1D,HIGH,BULLISH,,0.78,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,1555.258,1556.5,1560.23,1.197,0.694,-0.145,1.197,2.273,0.485,2.421,-1.348,2.421,-1.348,2.273,0.485,1,0
-2018-12-21,TCS.NS,3D,HIGH,BULLISH,,0.78,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,1555.258,1555.57,1558.06,1.197,0.694,-0.145,0.694,2.273,0.485,2.421,-1.348,2.421,-1.348,2.421,-1.348,1,1
-2018-12-21,TCS.NS,5D,HIGH,BULLISH,,0.78,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,1555.258,1555.57,1558.06,1.197,0.694,-0.145,-0.145,2.273,0.485,2.421,-1.348,2.421,-1.348,2.421,-1.348,1,1
-2018-12-21,WIPRO.NS,1D,HIGH,BULLISH,,0.68,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,110.585,110.67,110.94,1.428,1.661,2.717,1.428,2.608,-0.217,2.872,-0.854,3.632,-0.854,2.608,-0.217,1,1
-2018-12-21,WIPRO.NS,3D,HIGH,BULLISH,,0.68,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,110.585,110.61,110.78,1.428,1.661,2.717,1.661,2.608,-0.217,2.872,-0.854,3.632,-0.854,2.872,-0.854,1,1
-2018-12-21,WIPRO.NS,5D,HIGH,BULLISH,,0.68,16.0,True,github:gpt-4o-mini,github,gpt-4o-mini,110.585,110.61,110.78,1.428,1.661,2.717,2.717,2.608,-0.217,2.872,-0.854,3.632,-0.854,3.632,-0.854,1,1
-2019-02-19,BAJFINANCE.NS,1D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,249.736,249.94,250.54,1.811,3.149,3.826,1.811,2.149,0.438,4.461,0.438,4.618,0.438,2.149,0.438,1,0
-2019-02-19,BAJFINANCE.NS,3D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,249.736,249.79,250.19,1.811,3.149,3.826,3.149,2.149,0.438,4.461,0.438,4.618,0.438,4.461,0.438,1,0
-2019-02-19,BAJFINANCE.NS,5D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,249.736,249.79,250.19,1.811,3.149,3.826,3.826,2.149,0.438,4.461,0.438,4.618,0.438,4.618,0.438,1,0
-2019-02-19,HDFCBANK.NS,1D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,481.391,481.78,482.93,1.166,0.355,1.276,1.166,1.293,-0.103,2.114,-0.103,2.202,-0.103,1.293,-0.103,1,1
-2019-02-19,HDFCBANK.NS,3D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,481.391,481.49,482.26,1.166,0.355,1.276,0.355,1.293,-0.103,2.114,-0.103,2.202,-0.103,2.114,-0.103,1,1
-2019-02-19,HDFCBANK.NS,5D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,481.391,481.49,482.26,1.166,0.355,1.276,1.276,1.293,-0.103,2.114,-0.103,2.202,-0.103,2.202,-0.103,1,1
-2019-02-19,RELIANCE.NS,1D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,538.291,538.72,540.01,1.501,1.336,0.341,1.501,1.965,0.238,3.429,0.238,3.429,-0.831,1.965,0.238,1,0
-2019-02-19,RELIANCE.NS,3D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,538.291,538.4,539.26,1.501,1.336,0.341,1.336,1.965,0.238,3.429,0.238,3.429,-0.831,3.429,0.238,1,0
-2019-02-19,RELIANCE.NS,5D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,538.291,538.4,539.26,1.501,1.336,0.341,0.341,1.965,0.238,3.429,0.238,3.429,-0.831,3.429,-0.831,1,1
-2019-02-19,SUNPHARMA.NS,1D,MEDIUM,BULLISH,,0.57,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,384.269,384.58,385.5,2.103,4.073,5.343,2.103,2.611,0.58,5.584,0.58,6.37,0.58,2.611,0.58,1,0
-2019-02-19,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.57,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,384.269,384.35,384.96,2.103,4.073,5.343,4.073,2.611,0.58,5.584,0.58,6.37,0.58,5.584,0.58,1,0
-2019-02-19,SUNPHARMA.NS,5D,MEDIUM,BULLISH,,0.57,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,384.269,384.35,384.96,2.103,4.073,5.343,5.343,2.611,0.58,5.584,0.58,6.37,0.58,6.37,0.58,1,0
-2019-02-19,TCS.NS,1D,HIGH,BULLISH,,0.78,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,1565.991,1567.24,1571.0,0.522,1.095,7.03,0.522,1.522,-1.234,1.848,-1.234,7.368,-1.234,1.522,-1.234,1,1
-2019-02-19,TCS.NS,3D,HIGH,BULLISH,,0.78,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,1565.991,1566.3,1568.81,0.522,1.095,7.03,1.095,1.522,-1.234,1.848,-1.234,7.368,-1.234,1.848,-1.234,1,1
-2019-02-19,TCS.NS,5D,MEDIUM,BULLISH,,0.78,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,1565.991,1566.3,1568.81,0.522,1.095,7.03,7.03,1.522,-1.234,1.848,-1.234,7.368,-1.234,7.368,-1.234,1,1
-2019-02-19,WIPRO.NS,1D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,125.134,125.23,125.53,2.407,4.21,6.658,2.407,2.903,-0.619,4.554,-0.619,7.016,-0.619,2.903,-0.619,1,1
-2019-02-19,WIPRO.NS,3D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,125.134,125.16,125.36,2.407,4.21,6.658,4.21,2.903,-0.619,4.554,-0.619,7.016,-0.619,4.554,-0.619,1,1
-2019-02-19,WIPRO.NS,5D,MEDIUM,BULLISH,,0.64,18.5,False,github:gpt-4o-mini,github,gpt-4o-mini,125.134,125.16,125.36,2.407,4.21,6.658,6.658,2.903,-0.619,4.554,-0.619,7.016,-0.619,7.016,-0.619,1,1
-2019-04-23,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.71,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,296.102,296.34,297.05,1.911,2.036,3.186,1.911,2.107,-0.1,2.614,-0.1,3.753,-0.1,2.107,-0.1,1,1
-2019-04-23,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.71,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,296.102,296.16,296.63,1.911,2.036,3.186,2.036,2.107,-0.1,2.614,-0.1,3.753,-0.1,2.614,-0.1,1,1
-2019-04-23,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.71,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,296.102,296.16,296.63,1.911,2.036,3.186,3.186,2.107,-0.1,2.614,-0.1,3.753,-0.1,3.753,-0.1,1,1
-2019-04-23,HDFCBANK.NS,1D,HIGH,BULLISH,,0.64,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,518.638,519.05,520.3,1.534,1.588,4.921,1.534,1.719,0.078,2.383,0.078,5.093,0.078,1.719,0.078,1,1
-2019-04-23,HDFCBANK.NS,3D,HIGH,BULLISH,,0.64,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,518.638,518.74,519.57,1.534,1.588,4.921,1.588,1.719,0.078,2.383,0.078,5.093,0.078,2.383,0.078,1,1
-2019-04-23,HDFCBANK.NS,5D,HIGH,BULLISH,,0.64,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,518.638,518.74,519.57,1.534,1.588,4.921,4.921,1.719,0.078,2.383,0.078,5.093,0.078,5.093,0.078,1,1
-2019-04-23,RELIANCE.NS,1D,HIGH,BULLISH,,0.71,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,603.691,604.17,605.62,1.881,2.123,3.021,1.881,2.269,0.176,3.56,-0.092,3.67,-0.092,2.269,0.176,1,1
-2019-04-23,RELIANCE.NS,3D,HIGH,BULLISH,,0.71,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,603.691,603.81,604.78,1.881,2.123,3.021,2.123,2.269,0.176,3.56,-0.092,3.67,-0.092,3.56,-0.092,1,1
-2019-04-23,RELIANCE.NS,5D,HIGH,BULLISH,,0.71,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,603.691,603.81,604.78,1.881,2.123,3.021,3.021,2.269,0.176,3.56,-0.092,3.67,-0.092,3.67,-0.092,1,1
-2019-04-23,SUNPHARMA.NS,1D,HIGH,BULLISH,,0.68,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,435.27,435.62,436.66,0.171,-0.95,-3.489,0.171,0.309,-1.291,1.291,-1.931,1.291,-4.813,0.309,-1.291,1,1
-2019-04-23,SUNPHARMA.NS,3D,HIGH,BULLISH,,0.68,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,435.27,435.36,436.05,0.171,-0.95,-3.489,-0.95,0.309,-1.291,1.291,-1.931,1.291,-4.813,1.291,-1.931,1,1
-2019-04-23,SUNPHARMA.NS,5D,HIGH,BULLISH,,0.68,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,435.27,435.36,436.05,0.171,-0.95,-3.489,-3.489,0.309,-1.291,1.291,-1.931,1.291,-4.813,1.291,-4.813,1,1
-2019-04-23,TCS.NS,1D,HIGH,BULLISH,,0.78,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,1771.728,1773.15,1777.4,1.318,3.875,2.8,1.318,1.854,0.049,4.125,0.049,5.192,0.049,1.854,0.049,1,1
-2019-04-23,TCS.NS,3D,HIGH,BULLISH,,0.78,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,1771.728,1772.08,1774.92,1.318,3.875,2.8,3.875,1.854,0.049,4.125,0.049,5.192,0.049,4.125,0.049,1,1
-2019-04-23,TCS.NS,5D,HIGH,BULLISH,,0.78,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,1771.728,1772.08,1774.92,1.318,3.875,2.8,2.8,1.854,0.049,4.125,0.049,5.192,0.049,5.192,0.049,1,1
-2019-04-23,WIPRO.NS,1D,HIGH,BULLISH,,0.68,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,133.632,133.74,134.06,0.67,1.271,0.876,0.67,1.099,-0.309,1.907,-0.309,2.868,-0.309,1.099,-0.309,1,1
-2019-04-23,WIPRO.NS,3D,HIGH,BULLISH,,0.68,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,133.632,133.66,133.87,0.67,1.271,0.876,1.271,1.099,-0.309,1.907,-0.309,2.868,-0.309,1.907,-0.309,1,1
-2019-04-23,WIPRO.NS,5D,HIGH,BULLISH,,0.68,24.6,True,github:gpt-4o-mini,github,gpt-4o-mini,133.632,133.66,133.87,0.67,1.271,0.876,0.876,1.099,-0.309,1.907,-0.309,2.868,-0.309,2.868,-0.309,1,1
-2019-06-21,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.71,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,347.887,348.17,349.0,-0.575,1.742,3.232,-0.575,0.394,-0.92,1.939,-0.996,3.573,-0.996,0.394,-0.92,1,1
-2019-06-21,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.71,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,347.887,347.96,348.51,-0.575,1.742,3.232,1.742,0.394,-0.92,1.939,-0.996,3.573,-0.996,1.939,-0.996,1,1
-2019-06-21,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.71,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,347.887,347.96,348.51,-0.575,1.742,3.232,3.232,0.394,-0.92,1.939,-0.996,3.573,-0.996,3.573,-0.996,1,1
-2019-06-21,HDFCBANK.NS,1D,HIGH,BULLISH,,0.75,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,561.118,561.57,562.91,0.155,2.224,1.224,0.155,0.679,-0.174,2.311,-0.464,3.326,-0.464,0.679,-0.174,1,1
-2019-06-21,HDFCBANK.NS,3D,HIGH,BULLISH,,0.75,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,561.118,561.23,562.13,0.155,2.224,1.224,2.224,0.679,-0.174,2.311,-0.464,3.326,-0.464,2.311,-0.464,1,1
-2019-06-21,HDFCBANK.NS,5D,HIGH,BULLISH,,0.75,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,561.118,561.23,562.13,0.155,2.224,1.224,1.224,0.679,-0.174,2.311,-0.464,3.326,-0.464,3.326,-0.464,1,1
-2019-06-21,RELIANCE.NS,1D,MEDIUM,BULLISH,,0.61,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,566.354,566.81,568.17,-1.336,1.145,-2.063,-1.336,-0.238,-1.751,1.962,-1.973,1.962,-2.411,-0.238,-1.751,0,0
-2019-06-21,RELIANCE.NS,3D,HIGH,BULLISH,,0.61,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,566.354,566.47,567.37,-1.336,1.145,-2.063,1.145,-0.238,-1.751,1.962,-1.973,1.962,-2.411,1.962,-1.973,1,1
-2019-06-21,RELIANCE.NS,5D,HIGH,BULLISH,,0.61,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,566.354,566.47,567.37,-1.336,1.145,-2.063,-2.063,-0.238,-1.751,1.962,-1.973,1.962,-2.411,1.962,-2.411,1,1
-2019-06-21,SUNPHARMA.NS,1D,MEDIUM,BULLISH,,0.6,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,355.611,355.9,356.75,0.17,4.35,4.741,0.17,0.679,-1.385,5.421,-1.385,7.053,-1.385,0.679,-1.385,1,1
-2019-06-21,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.6,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,355.611,355.68,356.25,0.17,4.35,4.741,4.35,0.679,-1.385,5.421,-1.385,7.053,-1.385,5.421,-1.385,1,1
-2019-06-21,SUNPHARMA.NS,5D,MEDIUM,BULLISH,,0.6,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,355.611,355.68,356.25,0.17,4.35,4.741,4.741,0.679,-1.385,5.421,-1.385,7.053,-1.385,7.053,-1.385,1,1
-2019-06-21,TCS.NS,1D,HIGH,BULLISH,,0.8,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,1864.635,1866.13,1870.6,1.14,0.193,-1.007,1.14,1.34,0.042,1.34,0.007,1.34,-1.216,1.34,0.042,1,1
-2019-06-21,TCS.NS,3D,HIGH,BULLISH,,0.8,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,1864.635,1865.01,1867.99,1.14,0.193,-1.007,0.193,1.34,0.042,1.34,0.007,1.34,-1.216,1.34,0.007,1,1
-2019-06-21,TCS.NS,5D,HIGH,BULLISH,,0.8,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,1864.635,1865.01,1867.99,1.14,0.193,-1.007,-1.007,1.34,0.042,1.34,0.007,1.34,-1.216,1.34,-1.216,1,1
-2019-06-21,WIPRO.NS,1D,HIGH,BULLISH,,0.64,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,131.222,131.33,131.64,-0.682,0.105,-1.872,-0.682,0.612,-0.945,0.752,-1.399,0.752,-2.047,0.612,-0.945,1,1
-2019-06-21,WIPRO.NS,3D,HIGH,BULLISH,,0.64,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,131.222,131.25,131.46,-0.682,0.105,-1.872,0.105,0.612,-0.945,0.752,-1.399,0.752,-2.047,0.752,-1.399,1,1
-2019-06-21,WIPRO.NS,5D,HIGH,BULLISH,,0.64,14.6,True,github:gpt-4o-mini,github,gpt-4o-mini,131.222,131.25,131.46,-0.682,0.105,-1.872,-1.872,0.612,-0.945,0.752,-1.399,0.752,-2.047,0.752,-2.047,1,1
-2019-08-20,BAJFINANCE.NS,1D,MEDIUM,BULLISH,,0.64,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,321.938,322.2,322.97,-0.991,-3.635,1.932,-0.991,0.845,-1.454,0.845,-8.989,3.154,-8.989,0.845,-1.454,1,1
-2019-08-20,BAJFINANCE.NS,3D,MEDIUM,BULLISH,,0.64,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,321.938,322.0,322.52,-0.991,-3.635,1.932,-3.635,0.845,-1.454,0.845,-8.989,3.154,-8.989,0.845,-8.989,1,1
-2019-08-20,BAJFINANCE.NS,5D,MEDIUM,BULLISH,,0.64,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,321.938,322.0,322.52,-0.991,-3.635,1.932,1.932,0.845,-1.454,0.845,-8.989,3.154,-8.989,3.154,-8.989,1,1
-2019-08-20,HDFCBANK.NS,1D,MEDIUM,BULLISH,,0.64,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,517.269,517.68,518.92,0.236,-2.607,1.772,0.236,0.928,-0.288,0.928,-3.648,2.355,-3.648,0.928,-0.288,1,1
-2019-08-20,HDFCBANK.NS,3D,MEDIUM,BULLISH,,0.64,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,517.269,517.37,518.2,0.236,-2.607,1.772,-2.607,0.928,-0.288,0.928,-3.648,2.355,-3.648,0.928,-3.648,1,1
-2019-08-20,HDFCBANK.NS,5D,HIGH,BULLISH,,0.64,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,517.269,517.37,518.2,0.236,-2.607,1.772,1.772,0.928,-0.288,0.928,-3.648,2.355,-3.648,2.355,-3.648,1,1
-2019-08-20,RELIANCE.NS,1D,HIGH,BULLISH,,0.71,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,568.173,568.63,569.99,-0.392,-0.008,-0.086,-0.392,0.212,-0.741,0.631,-3.876,1.415,-3.876,0.212,-0.741,1,1
-2019-08-20,RELIANCE.NS,3D,MEDIUM,BULLISH,,0.71,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,568.173,568.29,569.2,-0.392,-0.008,-0.086,-0.008,0.212,-0.741,0.631,-3.876,1.415,-3.876,0.631,-3.876,1,1
-2019-08-20,RELIANCE.NS,5D,MEDIUM,BULLISH,,0.71,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,568.173,568.29,569.2,-0.392,-0.008,-0.086,-0.086,0.212,-0.741,0.631,-3.876,1.415,-3.876,1.415,-3.876,1,1
-2019-08-20,SUNPHARMA.NS,1D,MEDIUM,BULLISH,,0.64,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,393.545,393.86,394.8,-0.927,1.639,-0.986,-0.927,1.176,-1.734,2.15,-2.946,2.15,-2.946,1.176,-1.734,1,1
-2019-08-20,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.64,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,393.545,393.62,394.25,-0.927,1.639,-0.986,1.639,1.176,-1.734,2.15,-2.946,2.15,-2.946,2.15,-2.946,1,1
-2019-08-20,SUNPHARMA.NS,5D,MEDIUM,BULLISH,,0.64,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,393.545,393.62,394.25,-0.927,1.639,-0.986,-0.986,1.176,-1.734,2.15,-2.946,2.15,-2.946,2.15,-2.946,1,1
-2019-08-20,TCS.NS,1D,MEDIUM,BULLISH,,0.71,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,1816.571,1818.02,1822.38,-0.025,2.787,2.275,-0.025,0.697,-0.206,3.345,-0.766,4.356,-0.766,0.697,-0.206,1,1
-2019-08-20,TCS.NS,3D,MEDIUM,BULLISH,,0.71,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,1816.571,1816.93,1819.84,-0.025,2.787,2.275,2.787,0.697,-0.206,3.345,-0.766,4.356,-0.766,3.345,-0.766,1,1
-2019-08-20,TCS.NS,5D,MEDIUM,BULLISH,,0.71,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,1816.571,1816.93,1819.84,-0.025,2.787,2.275,2.275,0.697,-0.206,3.345,-0.766,4.356,-0.766,4.356,-0.766,1,1
-2019-08-20,WIPRO.NS,1D,MEDIUM,BULLISH,,0.71,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,116.28,116.37,116.65,-0.355,-0.75,-1.263,-0.355,0.632,-0.869,1.5,-1.895,1.5,-2.448,0.632,-0.869,1,1
-2019-08-20,WIPRO.NS,3D,MEDIUM,BULLISH,,0.71,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,116.28,116.3,116.49,-0.355,-0.75,-1.263,-0.75,0.632,-0.869,1.5,-1.895,1.5,-2.448,1.5,-1.895,1,1
-2019-08-20,WIPRO.NS,5D,MEDIUM,BULLISH,,0.71,16.6,False,github:gpt-4o-mini,github,gpt-4o-mini,116.28,116.3,116.49,-0.355,-0.75,-1.263,-1.263,0.632,-0.869,1.5,-1.895,1.5,-2.448,1.5,-2.448,1,1
-2019-10-22,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.75,16.8,True,github:gpt-4o-mini,github,gpt-4o-mini,394.264,394.58,395.53,0.104,-1.163,0.325,0.104,1.88,-1.64,1.88,-3.325,1.88,-3.325,1.88,-1.64,1,1
-2019-10-22,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.75,16.8,True,github:gpt-4o-mini,github,gpt-4o-mini,394.264,394.34,394.97,0.104,-1.163,0.325,-1.163,1.88,-1.64,1.88,-3.325,1.88,-3.325,1.88,-3.325,1,1
-2019-10-22,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.75,16.8,True,github:gpt-4o-mini,github,gpt-4o-mini,394.264,394.34,394.97,0.104,-1.163,0.325,0.325,1.88,-1.64,1.88,-3.325,1.88,-3.325,1.88,-3.325,1,1
-2019-12-18,HDFCBANK.NS,1D,HIGH,BULLISH,,0.8,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,602.083,602.56,604.01,-0.275,0.778,-1.695,-0.275,1.018,-0.484,1.018,-0.952,1.018,-2.143,1.018,-0.484,1,1
-2019-12-18,HDFCBANK.NS,3D,HIGH,BULLISH,,0.8,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,602.083,602.2,603.17,-0.275,0.778,-1.695,0.778,1.018,-0.484,1.018,-0.952,1.018,-2.143,1.018,-0.952,1,1
-2019-12-18,HDFCBANK.NS,5D,HIGH,BULLISH,,0.8,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,602.083,602.2,603.17,-0.275,0.778,-1.695,-1.695,1.018,-0.484,1.018,-0.952,1.018,-2.143,1.018,-2.143,1,1
-2019-12-18,RELIANCE.NS,1D,HIGH,BULLISH,,0.71,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,701.717,702.28,703.96,2.164,-0.282,-3.836,2.164,2.478,-0.257,2.646,-1.145,2.646,-4.169,2.478,-0.257,1,1
-2019-12-18,RELIANCE.NS,3D,HIGH,BULLISH,,0.71,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,701.717,701.86,702.98,2.164,-0.282,-3.836,-0.282,2.478,-0.257,2.646,-1.145,2.646,-4.169,2.646,-1.145,1,1
-2019-12-18,RELIANCE.NS,5D,HIGH,BULLISH,,0.71,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,701.717,701.86,702.98,2.164,-0.282,-3.836,-3.836,2.478,-0.257,2.646,-1.145,2.646,-4.169,2.646,-4.169,1,1
-2019-12-18,SUNPHARMA.NS,1D,HIGH,BULLISH,,0.64,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,411.17,411.5,412.49,-1.399,-2.581,-4.002,-1.399,0.193,-2.786,0.193,-2.99,0.193,-4.343,0.193,-2.786,1,0
-2019-12-18,SUNPHARMA.NS,3D,HIGH,BULLISH,,0.64,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,411.17,411.25,411.91,-1.399,-2.581,-4.002,-2.581,0.193,-2.786,0.193,-2.99,0.193,-4.343,0.193,-2.99,1,1
-2019-12-18,SUNPHARMA.NS,5D,HIGH,BULLISH,,0.64,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,411.17,411.25,411.91,-1.399,-2.581,-4.002,-4.002,0.193,-2.786,0.193,-2.99,0.193,-4.343,0.193,-4.343,1,1
-2019-12-18,TCS.NS,1D,HIGH,BULLISH,,0.75,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,1841.277,1842.75,1847.17,2.828,2.95,1.578,2.828,3.172,-0.009,3.642,-0.009,3.642,-0.009,3.172,-0.009,1,1
-2019-12-18,TCS.NS,3D,HIGH,BULLISH,,0.75,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,1841.277,1841.64,1844.59,2.828,2.95,1.578,2.95,3.172,-0.009,3.642,-0.009,3.642,-0.009,3.642,-0.009,1,1
-2019-12-18,TCS.NS,5D,HIGH,BULLISH,,0.75,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,1841.277,1841.64,1844.59,2.828,2.95,1.578,1.578,3.172,-0.009,3.642,-0.009,3.642,-0.009,3.642,-0.009,1,1
-2019-12-18,WIPRO.NS,1D,HIGH,BULLISH,,0.68,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,113.985,114.08,114.35,0.362,2.235,0.805,0.362,0.946,-0.262,2.578,-0.403,2.578,-0.403,0.946,-0.262,1,1
-2019-12-18,WIPRO.NS,3D,HIGH,BULLISH,,0.68,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,113.985,114.01,114.19,0.362,2.235,0.805,2.235,0.946,-0.262,2.578,-0.403,2.578,-0.403,2.578,-0.403,1,1
-2019-12-18,WIPRO.NS,5D,HIGH,BULLISH,,0.68,12.3,True,github:gpt-4o-mini,github,gpt-4o-mini,113.985,114.01,114.19,0.362,2.235,0.805,0.805,0.946,-0.262,2.578,-0.403,2.578,-0.403,2.578,-0.403,1,1
-2020-02-13,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.68,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,468.658,469.03,470.16,-0.284,-0.779,1.775,-0.284,0.41,-0.57,0.41,-1.955,2.67,-1.955,0.41,-0.57,1,1
-2020-02-13,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.68,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,468.658,468.75,469.5,-0.284,-0.779,1.775,-0.779,0.41,-0.57,0.41,-1.955,2.67,-1.955,0.41,-1.955,1,1
-2020-02-13,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.68,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,468.658,468.75,469.5,-0.284,-0.779,1.775,1.775,0.41,-0.57,0.41,-1.955,2.67,-1.955,2.67,-1.955,1,1
-2020-02-13,HDFCBANK.NS,1D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,578.346,578.81,580.2,-1.776,-2.268,-1.957,-1.776,0.564,-2.127,0.564,-3.053,0.564,-3.053,0.564,-2.127,1,1
-2020-02-13,HDFCBANK.NS,3D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,578.346,578.46,579.39,-1.776,-2.268,-1.957,-2.268,0.564,-2.127,0.564,-3.053,0.564,-3.053,0.564,-3.053,1,1
-2020-02-13,HDFCBANK.NS,5D,HIGH,BULLISH,,0.64,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,578.346,578.46,579.39,-1.776,-2.268,-1.957,-1.957,0.564,-2.127,0.564,-3.053,0.564,-3.053,0.564,-3.053,1,1
-2020-02-13,RELIANCE.NS,1D,HIGH,BULLISH,,0.71,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,656.43,656.96,658.53,0.912,-0.458,0.8,0.912,1.852,-1.428,2.171,-1.428,2.296,-1.428,1.852,-1.428,1,1
-2020-02-13,RELIANCE.NS,3D,HIGH,BULLISH,,0.71,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,656.43,656.56,657.61,0.912,-0.458,0.8,-0.458,1.852,-1.428,2.171,-1.428,2.296,-1.428,2.171,-1.428,1,1
-2020-02-13,RELIANCE.NS,5D,HIGH,BULLISH,,0.71,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,656.43,656.56,657.61,0.912,-0.458,0.8,0.8,1.852,-1.428,2.171,-1.428,2.296,-1.428,2.296,-1.428,1,1
-2020-02-13,SUNPHARMA.NS,1D,MEDIUM,BULLISH,,0.6,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,391.722,392.04,392.98,-0.095,-1.754,-2.644,-0.095,1.384,-0.585,1.384,-4.062,1.384,-4.062,1.384,-0.585,1,1
-2020-02-13,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.6,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,391.722,391.8,392.43,-0.095,-1.754,-2.644,-1.754,1.384,-0.585,1.384,-4.062,1.384,-4.062,1.384,-4.062,1,1
-2020-02-13,SUNPHARMA.NS,5D,MEDIUM,BULLISH,,0.6,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,391.722,391.8,392.43,-0.095,-1.754,-2.644,-2.644,1.384,-0.585,1.384,-4.062,1.384,-4.062,1.384,-4.062,1,1
-2020-02-13,TCS.NS,1D,HIGH,BULLISH,,0.78,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1866.06,1867.55,1872.03,-0.354,1.086,-1.604,-0.354,0.915,-0.956,1.325,-0.956,1.736,-1.857,0.915,-0.956,1,1
-2020-02-13,TCS.NS,3D,HIGH,BULLISH,,0.78,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1866.06,1866.43,1869.42,-0.354,1.086,-1.604,1.086,0.915,-0.956,1.325,-0.956,1.736,-1.857,1.325,-0.956,1,1
-2020-02-13,TCS.NS,5D,HIGH,BULLISH,,0.78,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1866.06,1866.43,1869.42,-0.354,1.086,-1.604,-1.604,0.915,-0.956,1.325,-0.956,1.736,-1.857,1.736,-1.857,1,1
-2020-02-13,WIPRO.NS,1D,MEDIUM,BULLISH,,0.57,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,112.348,112.44,112.71,-0.349,0.123,0.8,-0.349,0.656,-0.718,0.656,-1.026,1.969,-1.026,0.656,-0.718,1,1
-2020-02-13,WIPRO.NS,3D,MEDIUM,BULLISH,,0.57,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,112.348,112.37,112.55,-0.349,0.123,0.8,0.123,0.656,-0.718,0.656,-1.026,1.969,-1.026,0.656,-1.026,1,1
-2020-02-13,WIPRO.NS,5D,HIGH,BULLISH,,0.57,13.4,True,github:gpt-4o-mini,github,gpt-4o-mini,112.348,112.37,112.55,-0.349,0.123,0.8,0.8,0.656,-0.718,0.656,-1.026,1.969,-1.026,1.969,-1.026,1,1
-2020-04-17,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.71,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,226.098,226.28,226.82,0.004,-6.928,-14.379,0.004,1.618,-1.869,1.618,-12.382,1.618,-14.648,1.618,-1.869,1,1
-2020-04-17,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.71,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,226.098,226.14,226.5,0.004,-6.928,-14.379,-6.928,1.618,-1.869,1.618,-12.382,1.618,-14.648,1.618,-12.382,1,1
-2020-04-17,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.71,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,226.098,226.14,226.5,0.004,-6.928,-14.379,-14.379,1.618,-1.869,1.618,-12.382,1.618,-14.648,1.618,-14.648,1,1
-2020-04-17,HDFCBANK.NS,1D,MEDIUM,BULLISH,,0.57,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,424.093,424.43,425.45,3.795,2.01,3.048,3.795,5.564,2.713,5.564,-0.33,5.564,-0.33,5.564,2.713,1,0
-2020-04-17,HDFCBANK.NS,3D,MEDIUM,BULLISH,,0.57,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,424.093,424.18,424.86,3.795,2.01,3.048,2.01,5.564,2.713,5.564,-0.33,5.564,-0.33,5.564,-0.33,1,1
-2020-04-17,HDFCBANK.NS,5D,MEDIUM,BULLISH,,0.57,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,424.093,424.18,424.86,3.795,2.01,3.048,3.048,5.564,2.713,5.564,-0.33,5.564,-0.33,5.564,-0.33,1,1
-2020-04-17,RELIANCE.NS,1D,HIGH,BULLISH,,0.64,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,545.04,545.48,546.78,1.618,11.405,15.768,1.618,2.696,-1.797,13.145,-4.902,22.136,-4.902,2.696,-1.797,1,1
-2020-04-17,RELIANCE.NS,3D,HIGH,BULLISH,,0.64,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,545.04,545.15,546.02,1.618,11.405,15.768,11.405,2.696,-1.797,13.145,-4.902,22.136,-4.902,13.145,-4.902,1,1
-2020-04-17,RELIANCE.NS,5D,HIGH,BULLISH,,0.64,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,545.04,545.15,546.02,1.618,11.405,15.768,15.768,2.696,-1.797,13.145,-4.902,22.136,-4.902,22.136,-4.902,1,1
-2020-04-17,SUNPHARMA.NS,1D,MEDIUM,BULLISH,,0.61,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,430.478,430.82,431.86,3.566,3.774,6.224,3.566,3.916,-1.433,7.351,-1.433,8.729,-1.433,3.916,-1.433,1,1
-2020-04-17,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.61,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,430.478,430.56,431.25,3.566,3.774,6.224,3.774,3.916,-1.433,7.351,-1.433,8.729,-1.433,7.351,-1.433,1,1
-2020-04-17,SUNPHARMA.NS,5D,MEDIUM,BULLISH,,0.61,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,430.478,430.56,431.25,3.566,3.774,6.224,6.224,3.916,-1.433,7.351,-1.433,8.729,-1.433,8.729,-1.433,1,1
-2020-04-17,TCS.NS,1D,HIGH,BULLISH,,0.68,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,1548.896,1550.14,1553.85,0.689,-2.032,0.684,0.689,1.318,-0.205,1.318,-4.819,5.193,-4.819,1.318,-0.205,1,1
-2020-04-17,TCS.NS,3D,HIGH,BULLISH,,0.68,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,1548.896,1549.21,1551.68,0.689,-2.032,0.684,-2.032,1.318,-0.205,1.318,-4.819,5.193,-4.819,1.318,-4.819,1,1
-2020-04-17,TCS.NS,5D,HIGH,BULLISH,,0.68,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,1548.896,1549.21,1551.68,0.689,-2.032,0.684,0.684,1.318,-0.205,1.318,-4.819,5.193,-4.819,5.193,-4.819,1,1
-2020-04-17,WIPRO.NS,1D,MEDIUM,BULLISH,,0.57,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,86.191,86.26,86.47,-3.155,-4.519,-4.947,-3.155,1.07,-3.663,1.07,-7.059,1.07,-7.059,1.07,-3.663,1,1
-2020-04-17,WIPRO.NS,3D,MEDIUM,BULLISH,,0.57,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,86.191,86.21,86.35,-3.155,-4.519,-4.947,-4.519,1.07,-3.663,1.07,-7.059,1.07,-7.059,1.07,-7.059,1,1
-2020-04-17,WIPRO.NS,5D,MEDIUM,BULLISH,,0.57,42.6,False,github:gpt-4o-mini,github,gpt-4o-mini,86.191,86.21,86.35,-3.155,-4.519,-4.947,-4.947,1.07,-3.663,1.07,-7.059,1.07,-7.059,1.07,-7.059,1,1
-2020-06-16,BAJFINANCE.NS,1D,MEDIUM,BULLISH,,0.64,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,231.667,231.85,232.41,1.311,14.106,27.928,1.311,2.537,-1.393,14.78,-1.393,31.469,-1.393,2.537,-1.393,1,1
-2020-06-16,BAJFINANCE.NS,3D,MEDIUM,BULLISH,,0.64,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,231.667,231.71,232.08,1.311,14.106,27.928,14.106,2.537,-1.393,14.78,-1.393,31.469,-1.393,14.78,-1.393,1,1
-2020-06-16,BAJFINANCE.NS,5D,MEDIUM,BULLISH,,0.64,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,231.667,231.71,232.08,1.311,14.106,27.928,27.928,2.537,-1.393,14.78,-1.393,31.469,-1.393,31.469,-1.393,1,1
-2020-06-16,HDFCBANK.NS,1D,HIGH,BULLISH,,0.71,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,461.41,461.78,462.89,-1.126,4.337,5.24,-1.126,0.858,-1.732,5.311,-1.858,5.609,-1.858,0.858,-1.732,1,1
-2020-06-16,HDFCBANK.NS,3D,HIGH,BULLISH,,0.71,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,461.41,461.5,462.24,-1.126,4.337,5.24,4.337,0.858,-1.732,5.311,-1.858,5.609,-1.858,5.311,-1.858,1,1
-2020-06-16,HDFCBANK.NS,5D,HIGH,BULLISH,,0.71,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,461.41,461.5,462.24,-1.126,4.337,5.24,5.24,0.858,-1.732,5.311,-1.858,5.609,-1.858,5.609,-1.858,1,1
-2020-06-16,RELIANCE.NS,1D,HIGH,BULLISH,,0.72,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,727.181,727.76,729.51,-0.145,8.759,6.379,-0.145,1.097,-0.964,10.577,-0.964,11.529,-0.964,1.097,-0.964,1,1
-2020-06-16,RELIANCE.NS,3D,HIGH,BULLISH,,0.72,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,727.181,727.33,728.49,-0.145,8.759,6.379,8.759,1.097,-0.964,10.577,-0.964,11.529,-0.964,10.577,-0.964,1,1
-2020-06-16,RELIANCE.NS,5D,MEDIUM,BULLISH,,0.72,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,727.181,727.33,728.49,-0.145,8.759,6.379,6.379,1.097,-0.964,10.577,-0.964,11.529,-0.964,11.529,-0.964,1,1
-2020-06-16,SUNPHARMA.NS,1D,MEDIUM,BULLISH,,0.71,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,455.246,455.61,456.7,0.465,0.972,3.527,0.465,1.314,-0.89,1.769,-0.89,4.055,-0.89,1.314,-0.89,1,1
-2020-06-16,SUNPHARMA.NS,3D,MEDIUM,BULLISH,,0.71,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,455.246,455.34,456.07,0.465,0.972,3.527,0.972,1.314,-0.89,1.769,-0.89,4.055,-0.89,1.769,-0.89,1,1
-2020-06-16,SUNPHARMA.NS,5D,HIGH,BULLISH,,0.71,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,455.246,455.34,456.07,0.465,0.972,3.527,3.527,1.314,-0.89,1.769,-0.89,4.055,-0.89,4.055,-0.89,1,1
-2020-06-16,TCS.NS,1D,MEDIUM,BULLISH,,0.64,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,1759.521,1760.93,1765.15,0.098,-0.059,-0.513,0.098,0.638,-0.941,2.161,-1.007,2.161,-1.75,0.638,-0.941,1,1
-2020-06-16,TCS.NS,3D,MEDIUM,BULLISH,,0.64,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,1759.521,1759.87,1762.69,0.098,-0.059,-0.513,-0.059,0.638,-0.941,2.161,-1.007,2.161,-1.75,2.161,-1.007,1,1
-2020-06-16,TCS.NS,5D,MEDIUM,BULLISH,,0.64,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,1759.521,1759.87,1762.69,0.098,-0.059,-0.513,-0.513,0.638,-0.941,2.161,-1.007,2.161,-1.75,2.161,-1.75,1,1
-2020-06-16,WIPRO.NS,1D,MEDIUM,BULLISH,,0.64,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,97.898,97.98,98.21,2.542,4.543,3.861,2.542,2.99,-0.636,5.508,-0.636,5.508,-0.636,2.99,-0.636,1,1
-2020-06-16,WIPRO.NS,3D,MEDIUM,BULLISH,,0.64,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,97.898,97.92,98.07,2.542,4.543,3.861,4.543,2.99,-0.636,5.508,-0.636,5.508,-0.636,5.508,-0.636,1,1
-2020-06-16,WIPRO.NS,5D,MEDIUM,BULLISH,,0.64,33.0,False,github:gpt-4o-mini,github,gpt-4o-mini,97.898,97.92,98.07,2.542,4.543,3.861,3.861,2.99,-0.636,5.508,-0.636,5.508,-0.636,5.508,-0.636,1,1
-2020-08-11,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.64,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,338.674,338.95,339.76,-1.132,-3.653,-1.085,-1.132,-0.214,-2.759,0.422,-4.373,0.422,-4.373,-0.214,-2.759,0,0
-2020-08-11,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.64,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,338.674,338.74,339.28,-1.132,-3.653,-1.085,-3.653,-0.214,-2.759,0.422,-4.373,0.422,-4.373,0.422,-4.373,1,1
-2020-08-11,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.64,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,338.674,338.74,339.28,-1.132,-3.653,-1.085,-1.085,-0.214,-2.759,0.422,-4.373,0.422,-4.373,0.422,-4.373,1,1
-2020-08-11,HDFCBANK.NS,1D,HIGH,BULLISH,,0.64,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,496.933,497.33,498.52,-0.277,-3.019,-0.952,-0.277,-0.061,-1.748,0.441,-3.689,0.441,-4.355,-0.061,-1.748,0,0
-2020-08-11,HDFCBANK.NS,3D,HIGH,BULLISH,,0.64,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,496.933,497.03,497.83,-0.277,-3.019,-0.952,-3.019,-0.061,-1.748,0.441,-3.689,0.441,-4.355,0.441,-3.689,1,1
-2020-08-11,HDFCBANK.NS,5D,HIGH,BULLISH,,0.64,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,496.933,497.03,497.83,-0.277,-3.019,-0.952,-0.952,-0.061,-1.748,0.441,-3.689,0.441,-4.355,0.441,-4.355,1,1
-2020-08-11,RELIANCE.NS,1D,HIGH,BULLISH,,0.71,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,963.079,963.85,966.16,-0.291,-0.937,-0.715,-0.291,0.525,-1.265,1.087,-2.088,1.087,-2.985,0.525,-1.265,1,1
-2020-08-11,RELIANCE.NS,3D,HIGH,BULLISH,,0.71,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,963.079,963.27,964.81,-0.291,-0.937,-0.715,-0.937,0.525,-1.265,1.087,-2.088,1.087,-2.985,1.087,-2.088,1,1
-2020-08-11,RELIANCE.NS,5D,HIGH,BULLISH,,0.71,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,963.079,963.27,964.81,-0.291,-0.937,-0.715,-0.715,0.525,-1.265,1.087,-2.088,1.087,-2.985,1.087,-2.985,1,1
-2020-08-11,SUNPHARMA.NS,1D,HIGH,BULLISH,,0.68,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,509.538,509.95,511.17,-1.645,-1.793,-2.782,-1.645,-0.055,-2.089,-0.055,-4.066,-0.055,-4.066,-0.055,-2.089,0,0
-2020-08-11,SUNPHARMA.NS,3D,HIGH,BULLISH,,0.68,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,509.538,509.64,510.46,-1.645,-1.793,-2.782,-1.793,-0.055,-2.089,-0.055,-4.066,-0.055,-4.066,-0.055,-4.066,0,0
-2020-08-11,SUNPHARMA.NS,5D,HIGH,BULLISH,,0.68,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,509.538,509.64,510.46,-1.645,-1.793,-2.782,-2.782,-0.055,-2.089,-0.055,-4.066,-0.055,-4.066,-0.055,-4.066,0,0
-2020-08-11,TCS.NS,1D,HIGH,BULLISH,,0.71,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1965.261,1966.83,1971.55,-0.989,-1.656,-0.445,-0.989,0.724,-1.463,0.882,-2.053,0.882,-2.053,0.724,-1.463,1,1
-2020-08-11,TCS.NS,3D,HIGH,BULLISH,,0.71,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1965.261,1965.65,1968.8,-0.989,-1.656,-0.445,-1.656,0.724,-1.463,0.882,-2.053,0.882,-2.053,0.882,-2.053,1,1
-2020-08-11,TCS.NS,5D,HIGH,BULLISH,,0.71,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,1965.261,1965.65,1968.8,-0.989,-1.656,-0.445,-0.445,0.724,-1.463,0.882,-2.053,0.882,-2.053,0.882,-2.053,1,1
-2020-08-11,WIPRO.NS,1D,HIGH,BULLISH,,0.61,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,128.894,129.0,129.31,-1.234,-1.126,1.091,-1.234,0.375,-1.824,0.787,-1.824,2.771,-1.824,0.375,-1.824,1,1
-2020-08-11,WIPRO.NS,3D,HIGH,BULLISH,,0.61,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,128.894,128.92,129.13,-1.234,-1.126,1.091,-1.126,0.375,-1.824,0.787,-1.824,2.771,-1.824,0.787,-1.824,1,1
-2020-08-11,WIPRO.NS,5D,HIGH,BULLISH,,0.61,21.4,True,github:gpt-4o-mini,github,gpt-4o-mini,128.894,128.92,129.13,-1.234,-1.126,1.091,1.091,0.375,-1.824,0.787,-1.824,2.771,-1.824,2.771,-1.824,1,1
-2020-10-07,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.75,20.1,True,github:gpt-4o-mini,github,gpt-4o-mini,326.307,326.57,327.35,-0.141,-0.119,1.228,-0.141,1.707,-0.875,2.621,-0.875,2.621,-2.955,1.707,-0.875,1,1
-2020-10-07,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.75,20.1,True,github:gpt-4o-mini,github,gpt-4o-mini,326.307,326.37,326.89,-0.141,-0.119,1.228,-0.119,1.707,-0.875,2.621,-0.875,2.621,-2.955,2.621,-0.875,1,1
-2020-10-07,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.75,20.1,True,github:gpt-4o-mini,github,gpt-4o-mini,326.307,326.37,326.89,-0.141,-0.119,1.228,1.228,1.707,-0.875,2.621,-0.875,2.621,-2.955,2.621,-2.955,1,1
-2020-10-07,HDFCBANK.NS,1D,HIGH,BULLISH,,0.68,20.1,True,github:gpt-4o-mini,github,gpt-4o-mini,541.472,541.9,543.2,2.542,4.422,4.237,2.542,3.506,0.095,6.926,0.095,6.926,0.095,3.506,0.095,1,1
-2020-10-07,HDFCBANK.NS,3D,HIGH,BULLISH,,0.68,20.1,True,github:gpt-4o-mini,github,gpt-4o-mini,541.472,541.58,542.45,2.542,4.422,4.237,4.422,3.506,0.095,6.926,0.095,6.926,0.095,6.926,0.095,1,1
-2020-10-07,HDFCBANK.NS,5D,HIGH,BULLISH,,0.68,20.1,True,github:gpt-4o-mini,github,gpt-4o-mini,541.472,541.58,542.45,2.542,4.422,4.237,4.237,3.506,0.095,6.926,0.095,6.926,0.095,6.926,0.095,1,1
-2020-10-07,RELIANCE.NS,1D,HIGH,BULLISH,,0.75,20.1,True,github:gpt-4o-mini,github,gpt-4o-mini,1018.91,1019.73,1022.17,-0.808,-0.906,1.329,-0.808,0.465,-1.568,0.465,-1.79,2.06,-1.79,0.465,-1.568,1,1
-2020-10-07,RELIANCE.NS,3D,HIGH,BULLISH,,0.75,20.1,True,github:gpt-4o-mini,github,gpt-4o-mini,1018.91,1019.11,1020.74,-0.808,-0.906,1.329,-0.906,0.465,-1.568,0.465,-1.79,2.06,-1.79,0.465,-1.79,1,1
-2020-12-03,BAJFINANCE.NS,1D,HIGH,BULLISH,,0.68,19.0,True,github:gpt-4o-mini,github,gpt-4o-mini,476.798,477.18,478.32,0.162,-1.547,-1.493,0.162,1.86,-1.078,1.86,-1.899,1.86,-1.899,1.86,-1.078,1,1
-2020-12-03,BAJFINANCE.NS,3D,HIGH,BULLISH,,0.68,19.0,True,github:gpt-4o-mini,github,gpt-4o-mini,476.798,476.89,477.66,0.162,-1.547,-1.493,-1.547,1.86,-1.078,1.86,-1.899,1.86,-1.899,1.86,-1.899,1,1
-2020-12-03,BAJFINANCE.NS,5D,HIGH,BULLISH,,0.68,19.0,True,github:gpt-4o-mini,github,gpt-4o-mini,476.798,476.89,477.66,0.162,-1.547,-1.493,-1.493,1.86,-1.078,1.86,-1.899,1.86,-1.899,1.86,-1.899,1,1
-2020-12-03,HDFCBANK.NS,1D,HIGH,BULLISH,,0.75,19.0,True,github:gpt-4o-mini,github,gpt-4o-mini,641.613,642.13,643.67,0.61,-0.065,0.628,0.61,1.761,-0.283,1.761,-1.387,2.385,-1.387,1.761,-0.283,1,1
+2020-01-01,BAJFINANCE.NS,1D,MEDIUM,BULLISH,,0.78,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,413.532,413.74,415.39,0.349,-5.544,-4.286,0.349,1.523,0.087,1.523,-5.849,1.523,-6.648,1.523,0.087,1,1
+2020-01-01,BAJFINANCE.NS,3D,MEDIUM,BULLISH,,0.78,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,413.532,413.61,414.28,0.349,-5.544,-4.286,-5.544,1.523,0.087,1.523,-5.849,1.523,-6.648,1.523,-5.849,1,1
+2020-01-01,BAJFINANCE.NS,5D,MEDIUM,BULLISH,,0.78,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,413.532,413.61,414.28,0.349,-5.544,-4.286,-4.286,1.523,0.087,1.523,-5.849,1.523,-6.648,1.523,-6.648,1,1
+2020-01-01,DRREDDY.NS,1D,MEDIUM,NEUTRAL,,0.71,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,554.488,553.82,555.15,-0.504,-0.019,0.62,-0.504,0.451,-0.667,0.639,-0.95,1.063,-0.95,0.451,-0.667,1,1
+2020-01-01,DRREDDY.NS,3D,MEDIUM,NEUTRAL,,0.71,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,554.488,553.49,555.49,-0.504,-0.019,0.62,-0.019,0.451,-0.667,0.639,-0.95,1.063,-0.95,0.639,-0.95,1,1
+2020-01-01,DRREDDY.NS,5D,MEDIUM,NEUTRAL,,0.71,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,554.488,553.1,555.87,-0.504,-0.019,0.62,0.62,0.451,-0.667,0.639,-0.95,1.063,-0.95,1.063,-0.95,1,1
+2020-01-01,HDFCBANK.NS,1D,MEDIUM,NEUTRAL,,0.71,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,595.677,594.96,596.39,0.637,-2.945,-1.666,0.637,0.735,0.031,0.735,-3.332,0.735,-3.332,0.735,0.031,1,0
+2020-01-01,HDFCBANK.NS,3D,MEDIUM,NEUTRAL,,0.71,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,595.677,594.6,596.75,0.637,-2.945,-1.666,-2.945,0.735,0.031,0.735,-3.332,0.735,-3.332,0.735,-3.332,1,1
+2020-01-01,HDFCBANK.NS,5D,MEDIUM,NEUTRAL,,0.71,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,595.677,594.19,597.17,0.637,-2.945,-1.666,-1.666,0.735,0.031,0.735,-3.332,0.735,-3.332,0.735,-3.332,1,1
+2020-01-01,HINDUNILVR.NS,1D,MEDIUM,NEUTRAL,,0.67,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,1732.311,1730.23,1734.39,0.077,-1.09,-0.372,0.077,0.829,-0.338,0.829,-1.306,0.829,-1.554,0.829,-0.338,1,1
+2020-01-01,HINDUNILVR.NS,3D,MEDIUM,NEUTRAL,,0.67,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,1732.311,1729.19,1735.43,0.077,-1.09,-0.372,-1.09,0.829,-0.338,0.829,-1.306,0.829,-1.554,0.829,-1.306,1,1
+2020-01-01,HINDUNILVR.NS,5D,MEDIUM,NEUTRAL,,0.67,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,1732.311,1727.98,1736.64,0.077,-1.09,-0.372,-0.372,0.829,-0.338,0.829,-1.306,0.829,-1.554,0.829,-1.554,1,1
+2020-01-01,ICICIBANK.NS,1D,MEDIUM,NEUTRAL,,0.71,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,519.138,518.51,519.76,0.717,-2.059,-2.012,0.717,0.959,-0.168,0.959,-2.413,0.959,-4.052,0.959,-0.168,1,1
+2020-01-01,ICICIBANK.NS,3D,MEDIUM,NEUTRAL,,0.71,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,519.138,518.2,520.07,0.717,-2.059,-2.012,-2.059,0.959,-0.168,0.959,-2.413,0.959,-4.052,0.959,-2.413,1,1
+2020-01-01,ICICIBANK.NS,5D,MEDIUM,BULLISH,,0.71,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,519.138,519.24,520.07,0.717,-2.059,-2.012,-2.012,0.959,-0.168,0.959,-2.413,0.959,-4.052,0.959,-4.052,1,1
+2020-01-01,INFY.NS,1D,MEDIUM,BULLISH,,0.78,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,619.74,620.05,622.53,-0.292,0.271,-2.531,-0.292,0.536,-0.808,2.3,-0.808,2.3,-3.875,0.536,-0.808,1,1
+2020-01-01,INFY.NS,3D,MEDIUM,BULLISH,,0.78,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,619.74,619.86,620.86,-0.292,0.271,-2.531,0.271,0.536,-0.808,2.3,-0.808,2.3,-3.875,2.3,-0.808,1,1
+2020-01-01,INFY.NS,5D,HIGH,BULLISH,,0.78,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,619.74,619.86,620.86,-0.292,0.271,-2.531,-2.531,0.536,-0.808,2.3,-0.808,2.3,-3.875,2.3,-3.875,1,1
+2020-01-01,MARUTI.NS,1D,HIGH,BULLISH,,0.78,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,6945.336,6948.81,6976.59,0.248,-3.683,-3.782,0.248,0.77,0.004,0.77,-3.901,0.77,-4.53,0.77,0.004,1,1
+2020-01-01,MARUTI.NS,3D,MEDIUM,BULLISH,,0.78,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,6945.336,6946.73,6957.84,0.248,-3.683,-3.782,-3.683,0.77,0.004,0.77,-3.901,0.77,-4.53,0.77,-3.901,1,1
+2020-01-01,MARUTI.NS,5D,MEDIUM,BULLISH,,0.78,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,6945.336,6946.73,6957.84,0.248,-3.683,-3.782,-3.782,0.77,0.004,0.77,-3.901,0.77,-4.53,0.77,-4.53,1,1
+2020-01-01,NTPC.NS,1D,MEDIUM,NEUTRAL,,0.58,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,94.731,94.62,94.84,-0.123,-2.18,-1.316,-0.123,0.494,-0.782,0.494,-4.155,0.494,-4.155,0.494,-0.782,1,1
+2020-01-01,NTPC.NS,3D,MEDIUM,NEUTRAL,,0.58,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,94.731,94.56,94.9,-0.123,-2.18,-1.316,-2.18,0.494,-0.782,0.494,-4.155,0.494,-4.155,0.494,-4.155,1,1
+2020-01-01,NTPC.NS,5D,MEDIUM,BULLISH,,0.58,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,94.731,94.75,94.9,-0.123,-2.18,-1.316,-1.316,0.494,-0.782,0.494,-4.155,0.494,-4.155,0.494,-4.155,1,1
+2020-01-01,RELIANCE.NS,1D,MEDIUM,NEUTRAL,,0.64,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,672.216,671.41,673.02,1.702,-0.537,0.235,1.702,2.077,0.159,2.123,-0.768,2.123,-0.768,2.077,0.159,0,0
+2020-01-01,RELIANCE.NS,3D,MEDIUM,NEUTRAL,,0.64,11.7,True,github:gpt-4o-mini,github,gpt-4o-mini,672.216,671.01,673.43,1.702,-0.537,0.235,-0.537,2.077,0.159,2.123,-0.768,2.123,-0.768,2.123,-0.768,1,1

research/loop_backtest.py

@@ -145,125 +145,153 @@ def _replace_once(src: str, old: str, new: str) -> tuple[str, bool]:
     return src, False
 
 
-def _replace_in_synthesis_prompt(src: str, old: str, new: str) -> tuple[str, bool]:
-    start_marker = "def _build_synthesis_prompt("
-    end_marker = "def _downgrade_confidence("
-    s_idx = src.find(start_marker)
-    e_idx = src.find(end_marker)
-    if s_idx == -1 or e_idx == -1 or e_idx <= s_idx:
+def _find_in_synthesis(src: str, target: str) -> bool:
+    """Check if target string exists inside _build_synthesis_prompt function."""
+    start = src.find("def _build_synthesis_prompt(")
+    end = src.find("def _downgrade_confidence(", start)
+    if start == -1 or end == -1:
+        return False
+    return target in src[start:end]
+
+
+def _replace_in_synthesis(src: str, old: str, new: str) -> tuple[str, bool]:
+    """Replace string inside _build_synthesis_prompt only."""
+    start = src.find("def _build_synthesis_prompt(")
+    end = src.find("def _downgrade_confidence(", start)
+    if start == -1 or end == -1:
         return src, False
-    block = src[s_idx:e_idx]
+    block = src[start:end]
     if old not in block:
         return src, False
     block = block.replace(old, new, 1)
-    return src[:s_idx] + block + src[e_idx:], True
+    return src[:start] + block + src[end:], True
 
 
-def fix_widen_1d_bullish_range(src: str, iteration: int) -> tuple[str, str]:
-    """Widen 1D BULLISH hi so more upside moves are captured."""
+# ── PROMPT FIX FUNCTIONS ──────────────────────────────────────────────────────
+# All fixes target the signal alignment rules in _build_synthesis_prompt.
+# Strategy: the key lever for >90% accuracy is direction accuracy.
+# We do this by tightening the criteria that must be met before LLM calls directional.
+
+def fix_tighten_1d_bullish_rsi(src: str, iteration: int) -> tuple[str, str]:
+    """Tighten 1D BULLISH RSI threshold — require stronger oversold to call BULLISH."""
     pairs = [
-        ('  BULLISH: predicted_return_lo=0.08, predicted_return_hi=0.32  (midpoint≈0.20%)\n',
-         f'  BULLISH: predicted_return_lo=0.05, predicted_return_hi=0.45  (midpoint≈0.25%) [widen-1d-bull-{iteration}]\n'),
+        ("- BULLISH when: RSI < 40 (oversold bounce) OR (above EMA50 AND MACD > 0 AND volume high)\n",
+         f"- BULLISH when: RSI < 35 (deep oversold) OR (above EMA50 AND MACD > 0 AND volume > 1.3x avg) [v{iteration}]\n"),
+        (f"- BULLISH when: RSI < 35 (deep oversold) OR (above EMA50 AND MACD > 0 AND volume > 1.3x avg) [v{iteration-1}]\n",
+         f"- BULLISH when: RSI < 32 (extreme oversold) OR (above EMA50 AND MACD > 0 AND volume > 1.5x avg) [v{iteration}]\n"),
     ]
     for old, new in pairs:
-        new_src, ok = _replace_in_synthesis_prompt(src, old, new)
+        new_src, ok = _replace_in_synthesis(src, old, new)
         if ok:
-            return new_src, "  [FIX] 1D BULLISH hi widened to capture more upside"
-    return src, "  [FIX] 1D BULLISH anchor not found — skipped"
+            return new_src, f"  [FIX] 1D BULLISH RSI tightened to {32 if 'extreme' in new else 35}"
+    return src, "  [FIX] 1D BULLISH RSI anchor not found — skipped"
 
 
-def fix_tighten_1d_bearish(src: str, iteration: int) -> tuple[str, str]:
-    """Tighten 1D BEARISH magnitude — require stronger signal for BEARISH."""
+def fix_tighten_1d_bearish_rsi(src: str, iteration: int) -> tuple[str, str]:
+    """Tighten 1D BEARISH RSI threshold — require more overbought to call BEARISH."""
     pairs = [
-        ('  BEARISH: predicted_return_lo=-0.20, predicted_return_hi=-0.02  (midpoint≈-0.11%)\n',
-         f'  BEARISH: predicted_return_lo=-0.12, predicted_return_hi=-0.02  (midpoint≈-0.07%) [tight-1d-bear-{iteration}]\n'),
+        ("- BEARISH when: RSI > 68 AND below EMA50 AND MACD < 0 AND volume confirms\n",
+         f"- BEARISH when: RSI > 72 AND below EMA50 AND MACD < 0 AND volume confirms downside [v{iteration}]\n"),
+        (f"- BEARISH when: RSI > 72 AND below EMA50 AND MACD < 0 AND volume confirms downside [v{iteration-1}]\n",
+         f"- BEARISH when: RSI > 75 AND below BOTH EMA50 AND EMA200 AND MACD < 0 AND 90D return negative [v{iteration}]\n"),
     ]
     for old, new in pairs:
-        new_src, ok = _replace_in_synthesis_prompt(src, old, new)
+        new_src, ok = _replace_in_synthesis(src, old, new)
         if ok:
-            return new_src, "  [FIX] 1D BEARISH range tightened to reduce false negatives"
-    return src, "  [FIX] 1D BEARISH anchor not found — skipped"
-
-
-def fix_tighten_3d_bearish(src: str, iteration: int) -> tuple[str, str]:
-    """Tighten 3D BEARISH magnitude — shallower midpoint to avoid overshooting."""
-    # Target line index 1 (3D block) — need to replace only that instance
-    OLD = (
-        "- ACCURACY-CALIBRATED range values (use these as your starting point):\n"
-        '            "  BULLISH: predicted_return_lo=0.02, predicted_return_hi=0.18  (midpoint≈0.10%)\\n"\n'
-        '            "  BEARISH: predicted_return_lo=-0.18, predicted_return_hi=-0.02  (midpoint≈-0.10%)\\n"\n'
-        '            "  NEUTRAL: predicted_return_lo=-0.15, predicted_return_hi=0.15\\n"\n'
-    )
-    # Simpler approach: replace the specific 3D BEARISH line
-    old_bear = '  BEARISH: predicted_return_lo=-0.18, predicted_return_hi=-0.02  (midpoint≈-0.10%)\n'
-    new_bear = f'  BEARISH: predicted_return_lo=-0.12, predicted_return_hi=-0.02  (midpoint≈-0.07%) [tight-3d-bear-{iteration}]\n'
-    # The 3D BEARISH appears at index 1; replace only first occurrence AFTER 3D marker
-    start_marker = "TIMEFRAME CALIBRATION for 3D"
-    s_idx = src.find(start_marker)
-    if s_idx == -1:
-        return src, "  [FIX] 3D marker not found — skipped"
-    end_marker = "TIMEFRAME CALIBRATION for 5D"
-    e_idx = src.find(end_marker, s_idx)
-    if e_idx == -1:
-        e_idx = s_idx + 2000
-    block = src[s_idx:e_idx]
-    if old_bear not in block:
-        return src, "  [FIX] 3D BEARISH anchor not found — skipped"
-    block = block.replace(old_bear, new_bear, 1)
-    return src[:s_idx] + block + src[e_idx:], "  [FIX] 3D BEARISH range tightened"
-
-
-def fix_tighten_5d_bearish(src: str, iteration: int) -> tuple[str, str]:
-    """Tighten 5D BEARISH magnitude — shallower midpoint."""
-    old_bear = '  BEARISH: predicted_return_lo=-0.18, predicted_return_hi=-0.02  (midpoint≈-0.10%)\n'
-    new_bear = f'  BEARISH: predicted_return_lo=-0.12, predicted_return_hi=-0.02  (midpoint≈-0.07%) [tight-5d-bear-{iteration}]\n'
-    start_marker = "TIMEFRAME CALIBRATION for 5D"
-    s_idx = src.find(start_marker)
-    if s_idx == -1:
-        return src, "  [FIX] 5D marker not found — skipped"
-    block = src[s_idx:s_idx + 1500]
-    if old_bear not in block:
-        return src, "  [FIX] 5D BEARISH anchor not found — skipped"
-    block = block.replace(old_bear, new_bear, 1)
-    return src[:s_idx] + block + src[s_idx + 1500:], "  [FIX] 5D BEARISH range tightened"
-
-
-def fix_strengthen_directional_bias(src: str, iteration: int) -> tuple[str, str]:
-    """Strengthen the directional bias rule — raise the RSI threshold for BEARISH."""
+            return new_src, f"  [FIX] 1D BEARISH RSI threshold raised"
+    return src, "  [FIX] 1D BEARISH RSI anchor not found — skipped"
+
+
+def fix_tighten_3d_signal_count(src: str, iteration: int) -> tuple[str, str]:
+    """Tighten 3D direction criteria — require more evidence for directional calls."""
     pairs = [
-        ("- Predict BEARISH only when ALL of these apply: RSI > 62 AND price below EMA50 AND MACD bearish.\n",
-         f"- Predict BEARISH only when ALL of these apply: RSI > 65 AND price below BOTH EMA50 AND EMA200 AND MACD clearly bearish. [bias-strong-{iteration}]\n"),
+        ("- BEARISH when: below EMA50 AND (RSI > 58 OR MACD < 0) AND macro/sector headwinds\n",
+         f"- BEARISH when: below EMA50 AND RSI > 58 AND MACD < 0 AND macro/sector headwinds [v{iteration}]\n"),
+        (f"- BEARISH when: below EMA50 AND RSI > 58 AND MACD < 0 AND macro/sector headwinds [v{iteration-1}]\n",
+         f"- BEARISH when: below BOTH EMA50 AND EMA200 AND RSI > 60 AND MACD < 0 AND 90D return negative [v{iteration}]\n"),
     ]
     for old, new in pairs:
-        new_src, ok = _replace_in_synthesis_prompt(src, old, new)
+        new_src, ok = _replace_in_synthesis(src, old, new)
         if ok:
-            return new_src, "  [FIX] BEARISH RSI threshold raised to 65, requiring EMA200 confirmation"
-    return src, "  [FIX] directional bias anchor not found — skipped"
+            return new_src, "  [FIX] 3D BEARISH now requires both EMAs below and MACD confirmation"
+    return src, "  [FIX] 3D BEARISH anchor not found — skipped"
 
 
-def fix_widen_3d_5d_bullish(src: str, iteration: int) -> tuple[str, str]:
-    """Widen 3D/5D BULLISH hi slightly to capture more upside moves."""
-    old_bull = '  BULLISH: predicted_return_lo=0.02, predicted_return_hi=0.18  (midpoint≈0.10%)\n'
-    new_bull = f'  BULLISH: predicted_return_lo=0.02, predicted_return_hi=0.25  (midpoint≈0.13%) [widen-bull-{iteration}]\n'
-    new_src, ok = _replace_in_synthesis_prompt(src, old_bull, new_bull)
-    if ok:
-        return new_src, "  [FIX] 3D/5D BULLISH hi widened to 0.25"
-    return src, "  [FIX] 3D/5D BULLISH anchor not found — skipped"
+def fix_tighten_5d_direction(src: str, iteration: int) -> tuple[str, str]:
+    """Tighten 5D direction thresholds — only call directional when trend is clear."""
+    pairs = [
+        ("- NEUTRAL when: between EMAs, or any major signal is conflicting — prefer NEUTRAL over a weak guess\n",
+         f"- NEUTRAL when: between EMAs, OR RSI 40-62, OR MACD near zero, OR FII flows mixed — STRONGLY prefer NEUTRAL [v{iteration}]\n"),
+        (f"- NEUTRAL when: between EMAs, OR RSI 40-62, OR MACD near zero, OR FII flows mixed — STRONGLY prefer NEUTRAL [v{iteration-1}]\n",
+         f"- NEUTRAL when: any ambiguity at all in EMA position, RSI direction, or macro regime — NEUTRAL is correct answer [v{iteration}]\n"),
+    ]
+    for old, new in pairs:
+        new_src, ok = _replace_in_synthesis(src, old, new)
+        if ok:
+            return new_src, "  [FIX] 5D NEUTRAL threshold strengthened — more NEUTRAL calls"
+    return src, "  [FIX] 5D NEUTRAL anchor not found — skipped"
 
 
-# Fix sequence — ordered for target-hit (calibrated range fixes for 90% target)
+def fix_raise_vix_threshold(src: str, iteration: int) -> tuple[str, str]:
+    """Lower VIX bar for reducing BULLISH — now 18 instead of 20."""
+    pairs = [
+        ("- When VIX > 20 or macro is risk-off: require 4+ signals for BULLISH\n",
+         f"- When VIX > 18 or macro is risk-off: require 4+ signals for BULLISH; prefer NEUTRAL [v{iteration}]\n"),
+        (f"- When VIX > 18 or macro is risk-off: require 4+ signals for BULLISH; prefer NEUTRAL [v{iteration-1}]\n",
+         f"- When VIX > 16 or macro is risk-off: prefer NEUTRAL; only BULLISH with 5+ clear signals [v{iteration}]\n"),
+    ]
+    for old, new in pairs:
+        new_src, ok = _replace_in_synthesis(src, old, new)
+        if ok:
+            return new_src, "  [FIX] VIX BULLISH threshold lowered (18/16)"
+    return src, "  [FIX] VIX threshold anchor not found — skipped"
+
+
+def fix_increase_signal_count(src: str, iteration: int) -> tuple[str, str]:
+    """Require more signals to align before calling directional."""
+    pairs = [
+        ("- Require ≥3 of these to align before calling BULLISH or BEARISH:\n",
+         f"- Require ≥4 of these to align before calling BULLISH or BEARISH (≥3 is not enough): [v{iteration}]\n"),
+        (f"- Require ≥4 of these to align before calling BULLISH or BEARISH (≥3 is not enough): [v{iteration-1}]\n",
+         f"- Require ≥5 of these to clearly align before calling BULLISH or BEARISH: [v{iteration}]\n"),
+    ]
+    for old, new in pairs:
+        new_src, ok = _replace_in_synthesis(src, old, new)
+        if ok:
+            return new_src, "  [FIX] Required signal alignment count raised to 4/5"
+    return src, "  [FIX] signal count anchor not found — skipped"
+
+
+def fix_strengthen_neutral_preference(src: str, iteration: int) -> tuple[str, str]:
+    """Make NEUTRAL the strong default when signals conflict."""
+    pairs = [
+        ("- In genuine signal conflict: always choose NEUTRAL over a low-conviction directional call\n",
+         f"- RULE: When in doubt, output NEUTRAL. A wrong directional call is worse than NEUTRAL. [v{iteration}]\n"),
+        (f"- RULE: When in doubt, output NEUTRAL. A wrong directional call is worse than NEUTRAL. [v{iteration-1}]\n",
+         f"- RULE: NEUTRAL is the safe default. Only override to directional when evidence is overwhelming and specific. [v{iteration}]\n"),
+    ]
+    for old, new in pairs:
+        new_src, ok = _replace_in_synthesis(src, old, new)
+        if ok:
+            return new_src, "  [FIX] NEUTRAL preference strengthened in synthesis prompt"
+    return src, "  [FIX] NEUTRAL anchor not found — skipped"
+
+
+# Fix sequence — ordered for target-hit accuracy improvement
+# Each fix targets direction accuracy (the root cause of <90% target-hit)
 FIX_SEQUENCE = [
-    ("tighten_1d_bearish",          fix_tighten_1d_bearish),
-    ("tighten_3d_bearish",          fix_tighten_3d_bearish),
-    ("tighten_5d_bearish",          fix_tighten_5d_bearish),
-    ("strengthen_directional_bias", fix_strengthen_directional_bias),
-    ("widen_1d_bullish_range",      fix_widen_1d_bullish_range),
-    ("widen_3d_5d_bullish",         fix_widen_3d_5d_bullish),
+    ("tighten_5d_direction",          fix_tighten_5d_direction),
+    ("tighten_3d_signal_count",       fix_tighten_3d_signal_count),
+    ("tighten_1d_bullish_rsi",        fix_tighten_1d_bullish_rsi),
+    ("tighten_1d_bearish_rsi",        fix_tighten_1d_bearish_rsi),
+    ("raise_vix_threshold",           fix_raise_vix_threshold),
+    ("increase_signal_count",         fix_increase_signal_count),
+    ("strengthen_neutral_preference", fix_strengthen_neutral_preference),
 ]
 
 
 def choose_fix(res: dict, iteration: int) -> tuple[str, callable]:
-    """Pick fix based on target-hit accuracy (range prediction), not direction."""
+    """Pick fix based on which timeframe and direction has worst target-hit accuracy."""
     target_accs = [res[tf]["target_acc"] for tf in res if not np.isnan(res[tf].get("target_acc", float("nan")))]
     avg_target = np.mean(target_accs) if target_accs else float("nan")
 
@@ -275,26 +303,38 @@ def choose_fix(res: dict, iteration: int) -> tuple[str, callable]:
     t3 = r3d.get("target_acc", float("nan"))
     t5 = r5d.get("target_acc", float("nan"))
 
-    print(f"  Diagnosis: target_acc={avg_target:.1f}%  1D={t1:.1f}%  3D={t3:.1f}%  5D={t5:.1f}%")
+    b1 = r1d.get("dir_acc", float("nan"))   # direction accuracy 1D
+    b3 = r3d.get("dir_acc", float("nan"))   # direction accuracy 3D
+    b5 = r5d.get("dir_acc", float("nan"))   # direction accuracy 5D
+
+    print(f"  Diagnosis: target_acc={avg_target:.1f}%  1D={t1:.1f}% (dir={b1:.1f}%)  3D={t3:.1f}% (dir={b3:.1f}%)  5D={t5:.1f}% (dir={b5:.1f}%)")
 
-    # Prioritize weakest timeframe first (usually 5D).
+    # Identify weakest TF by target accuracy
     weakest = min(
         [("1D", t1), ("3D", t3), ("5D", t5)],
         key=lambda x: x[1] if not np.isnan(x[1]) else 999,
     )[0]
 
+    # 5D is hardest — fix its direction criteria first
     if weakest == "5D" and not np.isnan(t5) and t5 < TARGET:
-        return FIX_SEQUENCE[0]  # tighten_5d_range
+        return FIX_SEQUENCE[0]  # tighten_5d_direction
     if weakest == "3D" and not np.isnan(t3) and t3 < TARGET:
-        return FIX_SEQUENCE[1]  # tighten_3d_range
+        return FIX_SEQUENCE[1]  # tighten_3d_signal_count
     if weakest == "1D" and not np.isnan(t1) and t1 < TARGET:
-        return FIX_SEQUENCE[2]  # tighten_1d_range
-
-    # If all are present but still below target, enforce global clamps.
-    if not np.isnan(avg_target) and avg_target < TARGET - 15:
-        return FIX_SEQUENCE[3]  # enforce_atr_clamp
+        # Sub-diagnose: is BULLISH or BEARISH worse for 1D?
+        bull1 = r1d.get("target_bull", float("nan"))
+        bear1 = r1d.get("target_bear", float("nan"))
+        if not np.isnan(bull1) and not np.isnan(bear1) and bull1 < bear1:
+            return FIX_SEQUENCE[2]  # tighten_1d_bullish_rsi
+        return FIX_SEQUENCE[3]  # tighten_1d_bearish_rsi
+
+    # If all TFs present but still below target — try global fixes
+    if not np.isnan(avg_target) and avg_target < TARGET - 10:
+        return FIX_SEQUENCE[4]  # raise_vix_threshold
+    if not np.isnan(avg_target) and avg_target < TARGET - 5:
+        return FIX_SEQUENCE[5]  # increase_signal_count
     if not np.isnan(avg_target) and avg_target < TARGET:
-        return FIX_SEQUENCE[4]  # confidence_gap_tightness
+        return FIX_SEQUENCE[6]  # strengthen_neutral_preference
 
     return FIX_SEQUENCE[iteration % len(FIX_SEQUENCE)]
 
@@ -302,36 +342,27 @@ def choose_fix(res: dict, iteration: int) -> tuple[str, callable]:
 # ── MODEL STATUS ──────────────────────────────────────────────────────────────
 
 def _model_status() -> str:
-    """Return a one-line string showing which GitHub models are cooled down."""
+    """Return a one-line string showing which LLM providers are available."""
     try:
         import ai_forecast as _aif
-        _aif._load_model_cooldowns()
+        github_ready = bool(os.environ.get("GITHUB_TOKEN", ""))
+        or_ready = bool(os.environ.get("OPENROUTER_API_KEY", ""))
         parts = []
-        for m in _aif._GITHUB_MODEL_CANDIDATES:
-            cooled, remain = _aif._is_model_cooled_down(m)
-            if cooled:
-                parts.append(f"{m}: COOLDOWN {remain}s")
-            else:
-                parts.append(f"{m}: ready")
+        if github_ready:
+            parts.append("GitHub Models: ready")
+        if or_ready:
+            model = os.environ.get("OPENROUTER_BEST_FREE_MODEL", "openai/gpt-oss-120b:free")
+            parts.append(f"OpenRouter ({model}): ready")
+        if not parts:
+            parts.append("No LLM provider configured (set GITHUB_TOKEN or OPENROUTER_API_KEY)")
         return "  Models: " + " | ".join(parts)
     except Exception as e:
-        return f"  Models: (status unavailable: {e})"
+        return f"  Models: (status check failed: {e})"
 
 
 def _all_models_cooled_down() -> tuple[bool, int]:
-    """Return (all_cooled, max_wait_seconds)."""
-    try:
-        import ai_forecast as _aif
-        _aif._load_model_cooldowns()
-        max_wait = 0
-        for m in _aif._GITHUB_MODEL_CANDIDATES:
-            cooled, remain = _aif._is_model_cooled_down(m)
-            if not cooled:
-                return False, 0
-            max_wait = max(max_wait, remain)
-        return True, max_wait
-    except Exception:
-        return False, 0
+    """Return (all_cooled, max_wait_seconds) — always False unless we can detect cooldowns."""
+    return False, 0
 
 
 # ── MAIN LOOP ─────────────────────────────────────────────────────────────────

static/app.js

@@ -1826,6 +1826,7 @@ function _autoFillShares() {
 }
 
 let _pendingTradeData = {};
+let _pendingTradeContextPromise = null;
 
 function openTradeModal(ticker, name, price, stopLoss = 0, target = 0, planData = null) {
   _pendingTradeData = planData || {};
@@ -1861,7 +1862,7 @@ function openTradeModal(ticker, name, price, stopLoss = 0, target = 0, planData
   // If no prediction context came with planData, try the cached watchlist prediction.
   // Covers: shell cards clicked mid-load, and Portfolio "New Trade" with a typed ticker.
   if (ticker && !_pendingTradeData.prediction_data) {
-    fetch(`/api/watchlist-pick/${encodeURIComponent(ticker)}`)
+    _pendingTradeContextPromise = fetch(`/api/watchlist-pick/${encodeURIComponent(ticker)}`)
       .then(r => r.ok ? r.json() : null)
       .then(d => {
         if (!d || !d.pick) return;
@@ -1887,7 +1888,8 @@ function openTradeModal(ticker, name, price, stopLoss = 0, target = 0, planData
         if (tgtEl && (!tgtEl.value || parseFloat(tgtEl.value) === 0) && tf.expected_target_price) tgtEl.value = tf.expected_target_price;
         _autoFillShares();
       })
-      .catch(() => {}); // silent — prediction context is optional
+      .catch(() => {}) // silent — prediction context is optional
+      .finally(() => { _pendingTradeContextPromise = null; });
   }
 }
 
@@ -1916,6 +1918,15 @@ document.getElementById('modal-submit')?.addEventListener('click', async () => {
 
   if (!ticker || !dir || !entry || !shares) return alert('Ticker, direction, entry price, and shares are required.');
 
+  // Give the async context fetch a chance to complete before posting trade.
+  if (_pendingTradeContextPromise) {
+    try {
+      await _pendingTradeContextPromise;
+    } catch (_) {
+      // Best effort only; backend still auto-fills missing context.
+    }
+  }
+
   try {
     const res = await fetch('/api/trades', {
       method: 'POST',

top5_picker.py

@@ -116,16 +116,66 @@ def get_top5_picks(
             except Exception:
                 scan_preds[ticker] = {}
 
-    conf_order = {"HIGH": 0, "MEDIUM": 1, "LOW": 2, "BLOCKED": 9}
-    bullish = [
-        p for p in scan_preds.values()
-        if p
-        and p.get("direction") == "BULLISH"
-        and p.get("confidence") in ("HIGH", "MEDIUM", "LOW")
-        and p.get("no_trade_reason") != "ai_unavailable"
-    ]
-    bullish.sort(key=lambda x: (conf_order.get(x.get("confidence"), 5), -float(x.get("midpoint") or 0.0)))
-    candidates = bullish[:top_n]
+    # ── Composite 5D profit score ─────────────────────────────────────────────
+    # Goal: rank stocks by maximum expected profit over 5 days, not just confidence tier.
+    # Score components (all multiplicative on ret_hi so absolute return is preserved):
+    #   conf_mult:    HIGH=1.0 / MEDIUM=0.80 / LOW=0.55
+    #   ml_factor:    1 + (ml_probability - 0.5) × 0.30   → range [0.85, 1.15]
+    #   rr_factor:    1 + 0.12 if actual_rr >= 2.0 else 0  (rewards good risk/reward)
+    #   sector_factor: 1.12 if sector leading / 0.90 if sector lagging / 1.0 neutral
+    # Minimum thresholds: ret_hi > 1.0% AND (no R:R data OR actual_rr >= 1.2)
+    _CONF_MULT = {"HIGH": 1.0, "MEDIUM": 0.80, "LOW": 0.55}
+    _MIN_RET_HI = 1.0   # % — must have at least 1% upside headroom for a 5D hold
+    _MIN_RR     = 1.2   # minimum R:R when risk data is available
+
+    def _score_5d(p: dict) -> float:
+        """Composite profit score for a 5D scan result. Higher is better."""
+        ret_hi_val = float(p.get("ret_hi") or 0.0)
+        conf = p.get("confidence", "LOW")
+        conf_mult = _CONF_MULT.get(conf, 0.55)
+
+        ml_prob = float((p.get("ml") or {}).get("probability") or 0.5)
+        ml_factor = 1.0 + (ml_prob - 0.5) * 0.30
+
+        risk_data = p.get("risk") or {}
+        actual_rr = risk_data.get("actual_rr")
+        rr_factor = 1.12 if (actual_rr is not None and actual_rr >= 2.0) else 1.0
+
+        sector_data = p.get("sector") or {}
+        if sector_data.get("leading"):
+            sector_factor = 1.12
+        elif sector_data.get("lagging"):
+            sector_factor = 0.90
+        else:
+            sector_factor = 1.0
+
+        return ret_hi_val * conf_mult * ml_factor * rr_factor * sector_factor
+
+    # Filter: BULLISH, valid confidence, no ai_unavailable block, min thresholds
+    bullish = []
+    for p in scan_preds.values():
+        if not p:
+            continue
+        if p.get("direction") != "BULLISH":
+            continue
+        if p.get("confidence") not in ("HIGH", "MEDIUM", "LOW"):
+            continue
+        if p.get("no_trade_reason") == "ai_unavailable":
+            continue
+        ret_hi_val = float(p.get("ret_hi") or 0.0)
+        if ret_hi_val < _MIN_RET_HI:
+            continue
+        risk_data = p.get("risk") or {}
+        actual_rr = risk_data.get("actual_rr")
+        if actual_rr is not None and actual_rr < _MIN_RR:
+            continue
+        bullish.append(p)
+
+    # Sort by composite 5D profit score descending.
+    # Expand candidate pool to 2× top_n so the full 1D/3D rerun has room to re-rank.
+    bullish.sort(key=_score_5d, reverse=True)
+    candidate_pool_size = top_n * 2  # fetch 10 full predictions, trim to top_n at end
+    candidates = bullish[:candidate_pool_size]
 
     market_from_scan = next((p.get("market", {}) for p in scan_preds.values() if p and p.get("market")), {})
     shared_ctx = {
@@ -207,7 +257,9 @@ def get_top5_picks(
             "risk": {},
         }
 
-    # Step 3: Assemble picks
+    # Step 3: Assemble picks for full candidate pool, then re-rank by 5D profit score
+    # and trim to top_n. This ensures the final list maximises 5D return even after
+    # 1D/3D full predictions might have changed direction/confidence for some stocks.
     picks = []
     for stock in candidates:
         ticker  = stock["ticker"]
@@ -267,8 +319,20 @@ def get_top5_picks(
         pick["signals"] = {}
         pick["signal_count"] = 0
         pick["timeframes"] = timeframe_data
+
+        # Attach composite 5D score using the full (debate) prediction as anchor
+        pick["_score_5d"] = _score_5d(ai_anchor if (ai_anchor and not ai_anchor.get("error")) else stock)
         picks.append(pick)
 
+    # Re-rank by 5D composite profit score (uses full debate predictions where available)
+    picks.sort(key=lambda x: x.get("_score_5d", 0.0), reverse=True)
+    picks = picks[:top_n]
+
+    # Assign final ranks and remove internal score field
+    for i, p in enumerate(picks):
+        p["rank"] = i + 1
+        p.pop("_score_5d", None)
+
     return {
         "picks": picks,
         "market": market_from_scan,