diff --git a/backend/app/analysis/breakout_signals.py b/backend/app/analysis/breakout_signals.py
index fd4109c3..f4dcad25 100644
--- a/backend/app/analysis/breakout_signals.py
+++ b/backend/app/analysis/breakout_signals.py
@@ -196,14 +196,14 @@ def detect_pullback(df: pd.DataFrame) -> dict | None:
     2. 近 5 日内有 MA5 > MA20 — 近期处于上升趋势
     3. 价格靠近 MA10 或 MA20 (<3%)
     4. 近 3 日缩量（相对前 5 日）或 当日开始放量反弹
-    5. RSI 30-70（放宽）
+    5. RSI 只做风险提示，不作为回踩成立的硬条件
     """
     last = df.iloc[-1]
 
-    required = ["close", "vol", "ma5", "ma10", "ma20", "ma60", "rsi14"]
+    required = ["close", "vol", "ma5", "ma10", "ma20", "ma60"]
     if any(c not in df.columns for c in required):
         return None
-    if pd.isna(last["ma60"]) or pd.isna(last["rsi14"]):
+    if pd.isna(last["ma60"]):
         return None
 
     # 中长期上升趋势
@@ -251,13 +251,10 @@ def detect_pullback(df: pd.DataFrame) -> dict | None:
     if not (shrinking or bouncing_today):
         return None
 
-    # RSI（放宽至 30-70）
-    if not (30 <= last["rsi14"] <= 70):
-        return None
-
     support_ma = "MA10" if near_ma10 else "MA20"
     support_price = last["ma10"] if near_ma10 else last["ma20"]
     shrink_ratio = recent_3_vol / prev_5_vol if prev_5_vol > 0 else 1
+    rsi14 = last.get("rsi14") if "rsi14" in df.columns else None
 
     return {
         "signal_type": EntrySignal.PULLBACK,
@@ -266,6 +263,7 @@ def detect_pullback(df: pd.DataFrame) -> dict | None:
             "support_price": round(support_price, 2),
             "volume_shrink_ratio": round(shrink_ratio, 2),
             "bouncing": bouncing_today,
+            "rsi14": round(float(rsi14), 1) if rsi14 is not None and not pd.isna(rsi14) else None,
         },
     }
 
@@ -343,16 +341,16 @@ def detect_reversal(df: pd.DataFrame) -> dict | None:
     1. 近 5 日有 3 日以上下跌
     2. 今日放量长阳（pct_chg > 3%, vol > vol_ma5 * 1.5）
     3. 收盘价高于近 5 日最高价（强势反转）
-    4. RSI 从低位回升（近 5 日最低 RSI < 40）
+    4. RSI 低位只做背景信息，不作为反转成立的硬条件
     """
     if len(df) < 10:
         return None
 
     last = df.iloc[-1]
-    required = ["close", "high", "low", "vol", "pct_chg", "rsi14", "vol_ma5"]
+    required = ["close", "high", "low", "vol", "pct_chg", "vol_ma5"]
     if any(c not in df.columns for c in required):
         return None
-    if pd.isna(last["vol_ma5"]) or last["vol_ma5"] == 0 or pd.isna(last["rsi14"]):
+    if pd.isna(last["vol_ma5"]) or last["vol_ma5"] == 0:
         return None
 
     # 近 5 日有 3 日以上下跌
@@ -372,11 +370,10 @@ def detect_reversal(df: pd.DataFrame) -> dict | None:
     if last["close"] < high_5d:
         return None
 
-    # RSI 曾在低位
-    rsi_5d = df["rsi14"].tail(5)
-    min_rsi = rsi_5d.min()
-    if not pd.isna(min_rsi) and min_rsi > 40:
-        return None
+    min_rsi = None
+    if "rsi14" in df.columns:
+        rsi_5d = df["rsi14"].tail(5)
+        min_rsi = rsi_5d.min()
 
     return {
         "signal_type": EntrySignal.REVERSAL,
@@ -384,7 +381,7 @@ def detect_reversal(df: pd.DataFrame) -> dict | None:
             "down_days": down_count,
             "reversal_pct": round(last["pct_chg"], 2),
             "volume_ratio": round(last["vol"] / last["vol_ma5"], 2),
-            "min_rsi": round(float(min_rsi), 1),
+            "min_rsi": round(float(min_rsi), 1) if min_rsi is not None and not pd.isna(min_rsi) else None,
         },
     }
 
@@ -432,9 +429,9 @@ def _score_breakout(df: pd.DataFrame, signal: dict) -> float:
 
     if "rsi14" in df.columns and not pd.isna(last["rsi14"]):
         if 50 <= last["rsi14"] <= 75:
-            score += 10
+            score += 4
         elif 45 <= last["rsi14"] <= 80:
-            score += 5
+            score += 2
 
     return min(score, 100)
 
@@ -474,9 +471,9 @@ def _score_breakout_confirm(df: pd.DataFrame, signal: dict) -> float:
 
     if "rsi14" in df.columns and not pd.isna(last["rsi14"]):
         if 50 <= last["rsi14"] <= 70:
-            score += 15
+            score += 4
         elif 45 <= last["rsi14"] <= 75:
-            score += 8
+            score += 2
 
     return min(score, 100)
 
@@ -519,9 +516,9 @@ def _score_pullback(df: pd.DataFrame, signal: dict) -> float:
 
     if "rsi14" in df.columns and not pd.isna(last["rsi14"]):
         if 40 <= last["rsi14"] <= 55:
-            score += 15
+            score += 3
         elif 35 <= last["rsi14"] <= 65:
-            score += 8
+            score += 1
 
     return min(score, 100)
 
@@ -588,14 +585,6 @@ def _score_reversal(df: pd.DataFrame, signal: dict) -> float:
     elif vol_ratio > 1.5:
         score += 12
 
-    min_rsi = signal["details"].get("min_rsi", 50)
-    if min_rsi < 25:
-        score += 20  # 极度超卖反转
-    elif min_rsi < 30:
-        score += 15
-    elif min_rsi < 35:
-        score += 10
-
     # 站上 MA5 加分
     if "ma5" in df.columns and not pd.isna(last["ma5"]):
         if last["close"] > last["ma5"]:
diff --git a/backend/app/analysis/signals.py b/backend/app/analysis/signals.py
index 19787a59..193ee029 100644
--- a/backend/app/analysis/signals.py
+++ b/backend/app/analysis/signals.py
@@ -1,6 +1,6 @@
 """买卖信号生成
 
-基于技术指标判断 7 种买入信号，生成综合技术面评分。
+基于量价与趋势判断辅助信号。RSI 等滞后指标只做风险备注，不主导推荐。
 """
 
 import logging
@@ -53,7 +53,7 @@ def _check_macd_golden(df: pd.DataFrame) -> bool:
 
 
 def _check_rsi_healthy(df: pd.DataFrame) -> bool:
-    """RSI 健康区间: RSI(14) 在 40-70"""
+    """RSI 节奏区间: RSI(14) 在 40-70，仅作为辅助节奏提示"""
     if df.empty:
         return False
     rsi = df.iloc[-1].get("rsi14", 50)
@@ -258,7 +258,7 @@ def generate_signals(ts_code: str, name: str = "") -> TechnicalSignal:
         (ma_bullish, 15),
         (volume_breakout, 20),
         (macd_golden, 15),
-        (rsi_healthy, 10),
+        (rsi_healthy, 3),
         (pullback_support, 15),
         (big_yang, 15),
         (boll_support, 10),
diff --git a/backend/app/api/stocks.py b/backend/app/api/stocks.py
index 372e437d..57d235d4 100644
--- a/backend/app/api/stocks.py
+++ b/backend/app/api/stocks.py
@@ -384,12 +384,12 @@ async def diagnose_stock(ts_code: str, mode: str = Query("entry")):
         )
 
     signal_str = (
-        f"推荐体系评分: 趋势评分={signals.trend_score}/100（均线排列+高低点结构+MA20方向，主评分10%权重), "
-        f"辅助信号计数={signals.signal_count}/7（触发计分，仅供参考不参与主评分), "
+        f"推荐体系评分: 趋势评分={signals.trend_score}/100（均线排列+高低点结构+MA20方向，主评分辅助项), "
+        f"辅助信号计数={signals.signal_count}/7（触发计分，仅供节奏参考，不作为主评分裁判), "
         f"均线多头: {signals.ma_bullish}, "
         f"放量突破: {signals.volume_breakout}, "
         f"MACD金叉: {signals.macd_golden}, "
-        f"RSI健康: {signals.rsi_healthy}, "
+        f"RSI节奏区间: {signals.rsi_healthy}, "
         f"缩量回踩: {signals.pullback_support}, "
         f"放量长阳: {signals.big_yang}, "
         f"布林支撑: {signals.boll_support}, "
@@ -471,7 +471,7 @@ async def diagnose_stock(ts_code: str, mode: str = Query("entry")):
             rec_result = await db.execute(
                 text(
                     "SELECT score, supply_demand_score, price_action_score, "
-                    "technical_score, position_score, sector, signal "
+                    "capital_score, technical_score, position_score, sector, signal, entry_signal_type "
                     "FROM recommendations "
                     "WHERE ts_code = :code "
                     "AND (action_plan IN ('可操作', '重点关注') OR COALESCE(llm_score, 0) >= 6 OR score >= 56) "
@@ -484,12 +484,14 @@ async def diagnose_stock(ts_code: str, mode: str = Query("entry")):
                 rm = rec_row._mapping
                 rec_score_str = (
                     f"\n推荐体系评分: 综合={rm['score']}, "
-                    f"供需={rm['supply_demand_score']}(50%权重), "
-                    f"形态={rm['price_action_score']}(40%权重), "
-                    f"趋势={rm['technical_score']}(10%权重), "
+                    f"资金={rm['capital_score']}(资金顺势核心), "
+                    f"供需={rm['supply_demand_score']}, "
+                    f"形态={rm['price_action_score']}, "
+                    f"趋势={rm['technical_score']}, "
                     f"位置安全={rm['position_score']}, "
                     f"板块={rm['sector']}, "
-                    f"信号={rm['signal']}"
+                    f"信号={rm['signal']}, "
+                    f"入场类型={rm['entry_signal_type']}"
                 )
     except Exception:
         pass
@@ -556,11 +558,12 @@ async def diagnose_stock(ts_code: str, mode: str = Query("entry")):
 重要提示：
 1. 你不是在写传统研报，而是在给交易作战台输出结构化会诊意见。
 2. 如果有推荐体系评分、操作计划、跟踪信息，请优先沿用当前推荐体系，而不是另起一套标准。
-3. 趋势评分是推荐体系的技术面核心分数（均线排列40+高低点结构35+MA20方向25=满分100），辅助信号计数仅供参考不参与主评分。
-4. 位置安全评分高(>80)表示股价处于相对低位，低(<40)表示可能追高。
-5. 板块信息和推荐体系信息优先级高于单一技术指标。
-6. 先给结论和动作，再解释原因；不要先铺陈背景再拖到最后才下结论。
-7. 如果证据不足，也要明确给出“观察”或“回避”，不能写成模糊建议。
+3. 当前推荐体系以“资金顺势 + 主线板块 + 供需/量价 + 趋势”为主，技术指标只做节奏和风控确认。
+4. RSI、MACD、KDJ 等滞后指标不能单独决定买卖；RSI 超买只提示追高风险，超卖只提示弱势或反弹弹性，不等于可买或不可买。
+5. 位置安全评分高(>80)表示股价处于相对低位，低(<40)表示可能追高。
+6. 板块信息、资金面和推荐体系信息优先级高于单一技术指标。
+7. 先给结论和动作，再解释原因；不要先铺陈背景再拖到最后才下结论。
+8. 如果证据不足，也要明确给出“观察”或“回避”，不能写成模糊建议。
 {freshness_note}
 
 请严格按以下 Markdown 结构输出，不要写成泛泛长文：
diff --git a/backend/app/data/models.py b/backend/app/data/models.py
index b995b108..54510898 100644
--- a/backend/app/data/models.py
+++ b/backend/app/data/models.py
@@ -119,7 +119,7 @@ class TechnicalSignal(BaseModel):
     ma_bullish: bool = False        # 均线多头排列
     volume_breakout: bool = False   # 放量突破
     macd_golden: bool = False       # MACD金叉
-    rsi_healthy: bool = False       # RSI健康区间
+    rsi_healthy: bool = False       # RSI节奏区间（辅助参考）
     pullback_support: bool = False  # 缩量回踩支撑
     big_yang: bool = False          # 底部放量长阳
     boll_support: bool = False      # 布林带下轨支撑
@@ -148,8 +148,8 @@ class Recommendation(BaseModel):
     sector_score: float
     capital_score: float
     technical_score: float
-    supply_demand_score: float = 0  # 供需评分（主评分50%权重）
-    price_action_score: float = 0  # 价格行为评分（主评分40%权重）
+    supply_demand_score: float = 0  # 供需评分
+    price_action_score: float = 0  # 价格行为评分
     position_score: float = 50      # 位置安全得分
     valuation_score: float = 50     # 估值安全得分
     signal: str                     # BUY / SELL / HOLD
diff --git a/backend/app/engine/screener.py b/backend/app/engine/screener.py
index cf56b9c7..1c5ebdd9 100644
--- a/backend/app/engine/screener.py
+++ b/backend/app/engine/screener.py
@@ -3,10 +3,10 @@
 三阶段管道：
   Step 1: 主线定位 — 把实时板块/快照板块归一成系统 MarketTheme
   Step 2: 主题内选股 — 从主线主题成分、领涨股和实时异动中召回候选
-  Step 3: 深度分析 — 供需 + 价格行为 + 趋势 + LLM (10-15只推荐)
+  Step 3: 深度分析 — 资金顺势 + 供需 + 价格行为 + 趋势 + LLM
 
-评分公式：供需关系 50% + 价格行为 40% + 趋势 10%
-主题地位和资金流作为前置上下文，涨停/广度只作为辅助证据。
+评分公式：资金顺势 + 供需关系 + 价格行为 + 趋势。
+资金流和主题地位进入主评分，RSI/MACD 只作为节奏与风险参考。
 
 风险乘数：惩罚取最大而非叠加（防过度惩罚），奖励可叠加。
 
@@ -605,10 +605,12 @@ async def _build_recommendations(
     total = len(candidates)
     signal_counts = {"breakout": 0, "breakout_confirm": 0, "pullback": 0, "launch": 0, "reversal": 0, "none": 0}
     score_weights = strategy_profile.score_weights if strategy_profile else {
-        "supply_demand": 0.50,
-        "price_action": 0.40,
-        "trend": 0.10,
+        "capital_momentum": 0.30,
+        "supply_demand": 0.30,
+        "price_action": 0.25,
+        "trend": 0.15,
     }
+    score_weights = _normalize_score_weights(score_weights)
     signal_priority = strategy_profile.entry_signal_priority if strategy_profile else []
     buy_threshold = strategy_profile.buy_threshold if strategy_profile else 60
 
@@ -651,6 +653,8 @@ async def _build_recommendations(
             supply_demand_score = score_supply_demand(df)
             price_action_score = _score_price_action(df, entry_signal)
             trend_score = _score_trend(df)
+            capital_score = _score_capital_simple(stock)
+            flow_momentum_score = _score_flow_momentum(stock, sector, hot_sectors)
 
             last = df.iloc[-1]
             trend_penalty = 1.0
@@ -660,6 +664,7 @@ async def _build_recommendations(
                         trend_penalty = 0.82
 
             final_score = (
+                flow_momentum_score * score_weights["capital_momentum"] +
                 supply_demand_score * score_weights["supply_demand"] +
                 price_action_score * score_weights["price_action"] +
                 trend_score * score_weights["trend"]
@@ -699,6 +704,8 @@ async def _build_recommendations(
             if entry_signal.get("signal_score", 0) >= 80:
                 final_score *= 1.10
 
+            final_score *= _flow_confirmation_multiplier(stock, hot_theme_match, market_temp)
+
             if not hot_theme_match:
                 final_score *= 0.82
             elif hot_theme_match not in hot_sectors[:5]:
@@ -721,11 +728,25 @@ async def _build_recommendations(
             level = _score_to_level(final_score)
             signal = "HOLD"
             position_score = tech_signal.position_score if tech_signal else 50
+            flow_confirmed = _is_flow_confirmed(
+                stock=stock,
+                flow_momentum_score=flow_momentum_score,
+                trend_score=trend_score,
+                price_action_score=price_action_score,
+            )
+            effective_signal_name = signal_name
+            if signal_name == "none" and flow_confirmed:
+                effective_signal_name = "flow_momentum"
+
             if (
                 signal_type != EntrySignal.NONE
                 and entry_signal.get("signal_score", 0) >= 50
                 and position_score >= 30
                 and final_score >= buy_threshold
+            ) or (
+                flow_confirmed
+                and position_score >= 30
+                and final_score >= buy_threshold + 2
             ):
                 signal = "BUY"
 
@@ -796,13 +817,13 @@ async def _build_recommendations(
                 if target_price < entry_price * (1 + min_target_pct):
                     target_price = round(entry_price * (1 + min_target_pct), 2)
 
+            stock["entry_signal_type"] = effective_signal_name
             reasons = _generate_reasons(stock, entry_signal, tech_signal, df, intraday)
-            stock["entry_signal_type"] = signal_name
             risk_note = _generate_risk_note(market_temp, tech_signal, stock)
             vol_pattern = analyze_volume_pattern(df)
-            entry_timing = _generate_entry_timing(signal_name, intraday)
+            entry_timing = _generate_entry_timing(effective_signal_name, intraday)
             trade_plan = _build_trade_plan(
-                signal_type=signal_name,
+                signal_type=effective_signal_name,
                 score=final_score,
                 market_temp=market_temp,
                 sector_stage=sector_stage,
@@ -820,7 +841,7 @@ async def _build_recommendations(
                 score=round(final_score, 1),
                 market_temp_score=round(market_temp_score, 1),
                 sector_score=round(_get_sector_heat(sector, hot_sectors), 1),
-                capital_score=round(_score_capital_simple(stock), 1),
+                capital_score=round(capital_score, 1),
                 technical_score=round(trend_score, 1),
                 supply_demand_score=round(supply_demand_score, 1),
                 price_action_score=round(price_action_score, 1),
@@ -834,7 +855,7 @@ async def _build_recommendations(
                 risk_note=risk_note,
                 level=level,
                 strategy=strategy_profile.strategy_id if strategy_profile else "trend_breakout",
-                entry_signal_type=signal_name,
+                entry_signal_type=effective_signal_name,
                 entry_timing=entry_timing,
                 action_plan=trade_plan["action_plan"],
                 trigger_condition=trade_plan["trigger_condition"],
@@ -869,8 +890,9 @@ async def _build_recommendations(
                 "hot_theme_name": hot_theme_match.sector_name if hot_theme_match else "",
                 "hot_theme_aliases": hot_theme_match.theme_aliases if hot_theme_match else [],
                 "stock_role_hint": stock.get("stock_role_hint", "待判断"),
-                "entry_signal_type": signal_name,
+                "entry_signal_type": effective_signal_name,
                 "entry_signal_score": round(entry_signal.get("signal_score", 0), 1),
+                "flow_momentum_score": round(flow_momentum_score, 1),
                 "signal_matches_profile": signal_matches_profile,
                 "risk_tags": _build_risk_tags(market_temp, tech_signal, sector_stage, trend_penalty),
                 "focus_points": _build_focus_points(stock, entry_signal, tech_signal, vol_pattern, sector_stage),
@@ -1042,6 +1064,145 @@ async def _build_recommendations(
     return recommendations
 
 
+# ── 主评分辅助 ──
+
+
+def _normalize_score_weights(weights: dict[str, float]) -> dict[str, float]:
+    """兼容旧策略权重，并保证资金顺势进入主评分。"""
+    defaults = {
+        "capital_momentum": 0.30,
+        "supply_demand": 0.30,
+        "price_action": 0.25,
+        "trend": 0.15,
+    }
+    merged = {**defaults, **(weights or {})}
+    keys = ["capital_momentum", "supply_demand", "price_action", "trend"]
+    total = sum(max(float(merged.get(k, 0) or 0), 0) for k in keys)
+    if total <= 0:
+        return defaults
+    return {k: max(float(merged.get(k, 0) or 0), 0) / total for k in keys}
+
+
+def _score_flow_momentum(stock: dict, sector_name: str, hot_sectors: list[SectorInfo]) -> float:
+    """资金顺势评分：个股资金在场 + 主线板块顺风 + 活跃度确认。"""
+    main_net = float(stock.get("main_net_inflow", 0) or 0)
+    inflow_ratio = float(stock.get("inflow_ratio", 0) or 0)
+    turnover_rate = float(stock.get("turnover_rate", 0) or 0)
+    volume_ratio = stock.get("volume_ratio")
+    volume_ratio = float(volume_ratio) if volume_ratio not in (None, "") else 0.0
+    recall_score = float(stock.get("recall_score", 0) or 0)
+    sector_heat = _get_sector_heat(sector_name, hot_sectors)
+    sector_limit_up = _get_sector_limit_up(sector_name, hot_sectors)
+
+    score = 0.0
+
+    # 个股主力资金占 40 分。负流入不一票否决，但会明显降权。
+    if main_net > 15000:
+        score += 28
+    elif main_net > 8000:
+        score += 24
+    elif main_net > 3000:
+        score += 18
+    elif main_net > 0:
+        score += 10
+    elif main_net < -5000:
+        score -= 12
+    elif main_net < 0:
+        score -= 6
+
+    if inflow_ratio > 12:
+        score += 12
+    elif inflow_ratio > 8:
+        score += 9
+    elif inflow_ratio > 4:
+        score += 6
+    elif inflow_ratio > 0:
+        score += 3
+    elif inflow_ratio < -8:
+        score -= 8
+
+    # 主线板块和涨停广度占 25 分。
+    if sector_heat >= 80:
+        score += 16
+    elif sector_heat >= 65:
+        score += 12
+    elif sector_heat >= 50:
+        score += 8
+    elif sector_heat >= 35:
+        score += 4
+
+    if sector_limit_up >= 5:
+        score += 9
+    elif sector_limit_up >= 3:
+        score += 6
+    elif sector_limit_up >= 1:
+        score += 3
+
+    # 活跃度和召回强度占 35 分。
+    if volume_ratio >= 2.5:
+        score += 12
+    elif volume_ratio >= 1.8:
+        score += 9
+    elif volume_ratio >= 1.2:
+        score += 6
+    elif volume_ratio > 0:
+        score += 2
+
+    if 4 <= turnover_rate <= 12:
+        score += 10
+    elif 2 <= turnover_rate <= 20:
+        score += 7
+    elif turnover_rate > 0:
+        score += 3
+
+    score += min(max(recall_score, 0), 100) * 0.13
+
+    return round(max(0, min(score, 100)), 1)
+
+
+def _flow_confirmation_multiplier(stock: dict, hot_theme_match: SectorInfo | None, market_temp: MarketTemperature) -> float:
+    """资金与主线共振时加分，资金背离时降权。"""
+    main_net = float(stock.get("main_net_inflow", 0) or 0)
+    inflow_ratio = float(stock.get("inflow_ratio", 0) or 0)
+    volume_ratio = stock.get("volume_ratio")
+    volume_ratio = float(volume_ratio) if volume_ratio not in (None, "") else 0.0
+
+    multiplier = 1.0
+    if hot_theme_match and main_net > 3000 and inflow_ratio > 3:
+        multiplier += 0.05
+    if hot_theme_match and volume_ratio >= 1.5 and market_temp.temperature >= 50:
+        multiplier += 0.04
+    if main_net < -3000 and inflow_ratio < -3:
+        multiplier -= 0.10
+    elif main_net < 0 and not hot_theme_match:
+        multiplier -= 0.06
+    return max(0.82, min(multiplier, 1.12))
+
+
+def _is_flow_confirmed(
+    stock: dict,
+    flow_momentum_score: float,
+    trend_score: float,
+    price_action_score: float,
+) -> bool:
+    """允许资金顺势票进入候选，不必等 RSI/MACD 等滞后指标确认。"""
+    main_net = float(stock.get("main_net_inflow", 0) or 0)
+    inflow_ratio = float(stock.get("inflow_ratio", 0) or 0)
+    volume_ratio = stock.get("volume_ratio")
+    volume_ratio = float(volume_ratio) if volume_ratio not in (None, "") else 0.0
+    tags = set(stock.get("recall_tags", []) or [])
+
+    return (
+        flow_momentum_score >= 72
+        and main_net > 2000
+        and inflow_ratio > 2
+        and volume_ratio >= 1.15
+        and trend_score >= 45
+        and price_action_score >= 42
+        and bool(tags & {"hot_theme_core", "theme_leader", "top_theme_member", "sector_recall"})
+    )
+
+
 # ── 价格行为评分 ──
 
 
@@ -1309,6 +1470,7 @@ def _generate_entry_timing(signal_type: str, intraday: bool) -> str:
         "pullback": "盘中靠近支撑位时分批进场，尾盘14:30确认支撑有效可加仓",
         "launch": "早盘放量确认后即可进场，注意开盘9:30-10:00量能",
         "reversal": "午后13:30确认不回落再进场，避免早盘追高",
+        "flow_momentum": "优先看资金是否延续流入和板块前排是否继续走强，分时承接确认后再分批",
     }
     return timing_map.get(signal_type, "盘中观察量价配合，确认信号后进场")
 
@@ -1334,7 +1496,8 @@ def _build_trade_plan(
         "pullback": "回踩支撑",
         "launch": "缩量整理后启动",
         "reversal": "放量反转",
-    }.get(signal_type, "技术信号")
+        "flow_momentum": "资金顺势确认",
+    }.get(signal_type, "资金与量价信号")
 
     if market_temp.temperature < 40 or sector_stage in ("end",):
         action_plan = "观察"
@@ -1421,9 +1584,13 @@ def _generate_reasons(
                   EntrySignal.PULLBACK: "回踩型", EntrySignal.LAUNCH: "启动型",
                   EntrySignal.REVERSAL: "反转型"}
     entry_label = signal_map.get(signal_type, "")
+    if stock.get("entry_signal_type") == "flow_momentum":
+        entry_label = "资金顺势型"
 
     # 入场信号
-    if entry_label and signal_type:
+    if entry_label and stock.get("entry_signal_type") == "flow_momentum":
+        reasons.append("主线主题内资金顺势增强，优先跟踪承接而非等待滞后指标")
+    elif entry_label and signal_type:
         st = signal_type.value
         if st == "breakout":
             breakout_pct = details.get("breakout_pct", 0)
@@ -1492,6 +1659,8 @@ def _generate_risk_note(
         notes.append("启动型整理可能延长，注意时间成本")
     elif entry_type == "reversal":
         notes.append("反转型可能二次探底，确认底部后再加仓")
+    elif entry_type == "flow_momentum":
+        notes.append("资金顺势型需防板块分歧和资金一日游，重点看次日承接")
 
     if market.temperature < 30:
         notes.append("市场情绪偏冷，系统性风险较高")
@@ -1541,6 +1710,8 @@ def _build_focus_points(
     signal_type = entry_signal.get("signal_type")
     if signal_type and getattr(signal_type, "value", "none") != "none":
         points.append(f"确认{signal_type.value}信号是否延续")
+    elif stock.get("entry_signal_type") == "flow_momentum":
+        points.append("确认主力流入和板块前排强度是否延续")
     if stock.get("main_net_inflow", 0) > 0:
         points.append("观察主力流入是否继续放大")
     if vol_pattern.get("volume_trend"):
@@ -1623,7 +1794,7 @@ def _summarize_for_llm(df, entry_signal: dict, tech_signal: TechnicalSignal | No
                 vol_conclusion += "，近5日缩量"
         parts.append(vol_conclusion)
 
-    # ── MACD 结论 ──
+    # ── MACD 结论（节奏参考） ──
     dif = last.get("dif", 0) or 0
     dea = last.get("dea", 0) or 0
     macd_desc = ""
@@ -1643,19 +1814,19 @@ def _summarize_for_llm(df, entry_signal: dict, tech_signal: TechnicalSignal | No
             macd_desc += "，零轴上方(偏多)"
         else:
             macd_desc += "，零轴下方(偏空)"
-    parts.append(macd_desc or "MACD数据不足")
+    parts.append((macd_desc or "MACD数据不足") + "；仅作节奏参考")
 
-    # ── RSI 结论 ──
+    # ── RSI 结论（风险提示，不做买卖裁判） ──
     rsi = last.get("rsi14", 50)
     if not pd.isna(rsi):
         if rsi > 80:
-            parts.append(f"RSI14={rsi:.0f}，超买区，回调风险大")
+            parts.append(f"RSI14={rsi:.0f}，偏热，提示追高风险但不单独否决资金顺势")
         elif rsi > 70:
             parts.append(f"RSI14={rsi:.0f}，偏高，注意追高风险")
         elif rsi >= 40:
-            parts.append(f"RSI14={rsi:.0f}，健康区间")
+            parts.append(f"RSI14={rsi:.0f}，节奏中性")
         else:
-            parts.append(f"RSI14={rsi:.0f}，偏低，可能超卖")
+            parts.append(f"RSI14={rsi:.0f}，偏低，提示弱势或反弹弹性，不单独构成买点")
 
     # ── 价格位置结论 ──
     if tech_signal:
diff --git a/backend/app/llm/batch_screener.py b/backend/app/llm/batch_screener.py
index 1dcd9fa2..6718e021 100644
--- a/backend/app/llm/batch_screener.py
+++ b/backend/app/llm/batch_screener.py
@@ -27,6 +27,7 @@ async def prefilter_single_stock(candidate: dict, market_summary: str) -> dict:
 主题别名: {", ".join(candidate.get("hot_theme_aliases", []) or ["无"])}
 召回来源: {', '.join(candidate.get('recall_tags', []) or ['未标注'])}
 规则参考分: {candidate.get('quant_score', 0)}/100
+资金顺势分: {candidate.get('flow_momentum_score', 0)}/100
 位置安全: {candidate.get('position_score', 50)}/100
 当前价: {candidate.get('current_price', '未知')}
 主题阶段: {candidate.get('sector_stage', '未知')}
@@ -109,6 +110,7 @@ async def analyze_single_stock(candidate: dict, market_summary: str) -> dict:
 主题: {candidate.get('sector', '未知')}
 主线主题匹配: {"是，匹配 " + candidate.get("hot_theme_name", "") if candidate.get("hot_theme_matched") else "否"}
 规则参考分: {candidate.get('quant_score', 0)}/100
+资金顺势分: {candidate.get('flow_momentum_score', 0)}/100
 位置安全: {candidate.get('position_score', 50)}/100
 当前价: {candidate.get('current_price', '未知')}"""
 
diff --git a/backend/app/llm/prompts.py b/backend/app/llm/prompts.py
index b448e3c9..2269b38c 100644
--- a/backend/app/llm/prompts.py
+++ b/backend/app/llm/prompts.py
@@ -126,9 +126,11 @@ SINGLE_STOCK_ANALYSIS_PROMPT = """\
 
 你的原则：
 1. 量化分数只是参考，不是最终答案
-2. 如果板块地位、量价质量、位置或时机不匹配，可以直接否决高分股
-3. 如果股票具备明确触发与失效边界，即使量化分不是最高，也可以提升优先级
-4. 输出的是交易裁决单，不是研报
+2. A股先看资金是否在主线里形成合力，再看K线形态是否给出可执行边界
+3. RSI、MACD、KDJ 这类滞后指标只能作为节奏和风险备注，不能因为超买/超卖本身否决资金顺势标的
+4. 如果板块地位、资金延续、量价质量、位置或时机不匹配，可以直接否决高分股
+5. 如果股票具备明确触发与失效边界，即使量化分不是最高，也可以提升优先级
+6. 输出的是交易裁决单，不是研报
 
 请严格输出 JSON，不要输出 Markdown，不要添加多余解释。字段如下：
 {
@@ -163,9 +165,10 @@ STOCK_PREFILTER_PROMPT = """\
 
 你的原则：
 1. 这一步不是最终买卖结论，只做资源分配
-2. 不能因为某一个规则分数低就直接忽略，要看题材位置、角色、量价异常、时机感
-3. 可以容忍不标准的形态，但不能容忍明显失真、明显追高、明显没有交易边界的票
-4. 输出必须是 JSON，不要输出 Markdown
+2. 优先识别资金正在进攻的主线前排或强承接标的
+3. 不能因为 RSI/MACD/KDJ 等滞后指标“不好看”就直接忽略，要看题材位置、角色、量价异常、时机感
+4. 可以容忍不标准的形态，但不能容忍明显失真、明显追高、明显没有交易边界的票
+5. 输出必须是 JSON，不要输出 Markdown
 
 字段格式：
 {
diff --git a/backend/app/llm/strategy_selector.py b/backend/app/llm/strategy_selector.py
index 8b304033..c305448f 100644
--- a/backend/app/llm/strategy_selector.py
+++ b/backend/app/llm/strategy_selector.py
@@ -50,7 +50,7 @@ def get_strategy_profile_by_id(strategy_id: str) -> StrategyProfile:
             name="主线突破",
             description="市场偏强，优先寻找主线板块内的突破和突破确认。",
             entry_signal_priority=["breakout", "breakout_confirm", "launch", "pullback", "reversal"],
-            score_weights={"supply_demand": 0.45, "price_action": 0.35, "trend": 0.20},
+            score_weights={"capital_momentum": 0.30, "supply_demand": 0.30, "price_action": 0.25, "trend": 0.15},
             min_score=62,
             buy_threshold=66,
             max_position_pct=30,
@@ -67,7 +67,7 @@ def get_strategy_profile_by_id(strategy_id: str) -> StrategyProfile:
             name="回踩轮动",
             description="市场震荡分化，优先做回踩支撑和板块轮动中的低吸确认。",
             entry_signal_priority=["pullback", "breakout_confirm", "launch", "breakout", "reversal"],
-            score_weights={"supply_demand": 0.40, "price_action": 0.30, "trend": 0.30},
+            score_weights={"capital_momentum": 0.28, "supply_demand": 0.30, "price_action": 0.22, "trend": 0.20},
             min_score=60,
             buy_threshold=63,
             max_position_pct=20,
@@ -84,7 +84,7 @@ def get_strategy_profile_by_id(strategy_id: str) -> StrategyProfile:
             name="启动试错",
             description="市场偏弱，适合少量观察启动型和反转型机会，不做强追涨。",
             entry_signal_priority=["launch", "reversal", "pullback", "breakout_confirm", "breakout"],
-            score_weights={"supply_demand": 0.35, "price_action": 0.35, "trend": 0.30},
+            score_weights={"capital_momentum": 0.32, "supply_demand": 0.28, "price_action": 0.25, "trend": 0.15},
             min_score=58,
             buy_threshold=61,
             max_position_pct=10,
@@ -101,7 +101,7 @@ def get_strategy_profile_by_id(strategy_id: str) -> StrategyProfile:
             name="防守观察",
             description="市场退潮，系统以观察池为主，不主动扩大出手。",
             entry_signal_priority=["pullback", "launch", "reversal", "breakout_confirm", "breakout"],
-            score_weights={"supply_demand": 0.35, "price_action": 0.40, "trend": 0.25},
+            score_weights={"capital_momentum": 0.35, "supply_demand": 0.25, "price_action": 0.25, "trend": 0.15},
             min_score=56,
             buy_threshold=64,
             max_position_pct=5,