1

AGENTS.md

@@ -75,7 +75,7 @@ AlphaX 是一个以 `Python + FastAPI + PostgreSQL + Docker + 静态 HTML` 组
 - `app/cli.py`
   - 统一命令入口：`screener`, `confirm`, `tracker`, `paper-trader`, `price-streamer`, `market`, `review`, `event`, `sentiment`, `onchain`, `llm-insights`。
 
-## 4. 代码主线
+## 4. 代码结构
 
 ### 4.1 推荐系统业务闭环
 
@@ -88,7 +88,7 @@ AlphaX 是一个以 `Python + FastAPI + PostgreSQL + Docker + 静态 HTML` 组
 3. `app/services/altcoin_confirm.py`
    负责确认，判断候选是否形成更可执行的机会，并生成入场计划、上下文和推送候选。
 4. `app/services/event_driven_screener.py`
-   负责事件/舆情驱动的快速触发检查，是技术筛选主链路的补充入口。
+   负责事件/舆情驱动的快速触发检查，是多策略发现层的补充入口。
 5. `app/services/price_streamer.py`
    负责实时价格缓存，不等同于完整推荐状态跟踪。
 6. `app/services/price_tracker.py`
@@ -104,7 +104,7 @@ AlphaX 是一个以 `Python + FastAPI + PostgreSQL + Docker + 静态 HTML` 组
 
 - 核心认知：因子不等于策略。一个因子可以是先决条件、触发、确认、入场、风控或归因，但不能因为单个因子表现好就直接升级成完整策略。
 - 完整策略必须至少包含：适用市场环境、交易宇宙、先决条件、核心触发、辅助确认、入场规则、止盈止损、失效条件、仓位/杠杆约束和独立复盘口径。
-- 多策略链路必须保持独立：`main_composite_v1`、`box_retest_1h_v1`、`box_retest_4h_v1` 等策略可以共享行情、账户级风控和执行框架，但不能共享同一套入场门槛、RR、挂单距离和 paper trading 执行门禁。
+- 多策略链路必须保持独立：`main_composite_v1`、`box_retest_1h_v1`、`box_retest_4h_v1`、`volume_ignition_1h_v1`、`compression_breakout_4h_v1`、`intraday_momentum_15m_v1` 等策略可以共享行情、账户级风控和执行框架，但不能共享同一套入场门槛、RR、挂单距离和 paper trading 执行门禁。
 - 策略级配置入口在 `app/core/strategy_registry.py`，`StrategyDefinition.entry_gate_config` 控制确认/跟踪/展示派生的买点质量闸门，`StrategyDefinition.paper_config` 控制该策略进入 paper trading 的入场、挂单和动态杠杆门槛。
 - 新增策略时必须注册稳定 `strategy_code`，并明确自己的 `entry_gate_config` / `paper_config`。不要把新策略的特殊门槛写进全局 `paper_trading_config()` 或 `DEFAULT_ENTRY_GATE`，否则会影响其他策略的信号生成和成交样本。
 - `apply_entry_quality_gate()` 必须传入或从 `entry_plan.strategy_code` 派生策略身份；`paper_trader.py` 中开仓、挂单、挂单成交和挂单维护应通过策略级配置合并后的参数执行。
@@ -126,8 +126,8 @@ AlphaX 是一个以 `Python + FastAPI + PostgreSQL + Docker + 静态 HTML` 组
 - 多策略改造计划记录在 `docs/MULTI_STRATEGY_ARCHITECTURE.md`。后续做策略级改造前必须先阅读并更新该文档。
 - 目标架构是：统一交易宇宙 -> 多个独立策略并行扫描 -> 标准策略信号 -> 冲突/重复仲裁 -> 推荐/观察/挂单 -> paper trading 保留策略血缘 -> 按策略独立复盘。
 - `strategy_version` 只表示版本，不应替代策略身份；后续推荐、挂单和交易账本都应补充 `strategy_code`、`strategy_signal_id`、`strategy_snapshot_json` 和 `factor_roles_json`。
-- 现有综合确认链路在迁移期应标记为 `main_composite_v1`，避免无策略来源的推荐继续进入 paper trading。
+- 现有综合确认策略在迁移期标记为 `main_composite_v1`，它只是平等策略之一，用于避免无策略来源的推荐继续进入 paper trading。
-- 第一个建议拆出的独立策略是 `box_retest_4h_v1`：核心触发来自 `box_breakout_pullback_4h`，但策略成立还需要市场环境、交易量、回踩距离、15m/1H 承接、盈亏比、失效条件和账户风控。
+- 当前已拆出的独立策略包括：`box_retest_1h_v1` / `box_retest_4h_v1` 箱体突破回踩、`volume_ignition_1h_v1` 1H放量突破启动、`compression_breakout_4h_v1` 4H压缩蓄力突破、`intraday_momentum_15m_v1` 15m日内动量延续。它们可以共享交易宇宙和行情数据，但必须保留各自的触发、入场、失效和 paper trading 门禁。
 - 新增策略必须先 observe-only 或 paper-only 积累样本，再进入灰度/发布；不能因为某个因子短期表现好就直接同步真实交易。
 
 ### 4.1.3 链上数据源
@@ -180,7 +180,7 @@ AlphaX 是一个以 `Python + FastAPI + PostgreSQL + Docker + 静态 HTML` 组
   - 市场环境识别中心，第一版基于市场快照、BTC/ETH 涨跌、山寨涨跌广度、强势/大跌数量和 funding 热度识别 `risk_off`、`btc_main_uptrend`、`altcoin_rotation`、`sideways_chop`、`meme_frenzy`、`unknown`。
 - `app/core/global_risk.py`
   - paper trading 全局风控门禁。单币机会进入开仓或挂单成交前，需要先检查市场环境和账户风险；critical 禁止新开仓，high 只允许高质量机会。
-- 多策略基础设施当前内置 `main_composite_v1`、`box_retest_4h_v1`、`box_retest_1h_v1`。`box_breakout_pullback_4h` / `box_breakout_pullback_1h` 只是触发因子，只有和入场确认、风控、失效条件组成完整剧本后，才作为对应策略信号写入 `strategy_signals`。
+- 多策略基础设施当前内置 `main_composite_v1`、`box_retest_4h_v1`、`box_retest_1h_v1`、`volume_ignition_1h_v1`、`compression_breakout_4h_v1`、`intraday_momentum_15m_v1`。`box_breakout_pullback_4h` / `box_breakout_pullback_1h`、`vp_fly_1h_current`、`short_tf_15m_ignition` 等只是因子，只有和入场确认、风控、失效条件组成完整剧本后，才作为对应策略信号写入 `strategy_signals`。
   - 确认层也会应用同一市场风控语义：`risk_level=critical` 且 `position_multiplier=0` 时，强势发现仍可记录为观察，但不能输出 `buy_now` 或新挂单动作；已有活跃可交易推荐会被降级为观察并写入 `market_risk_gate`。
 
 ## 5. 数据与状态中心
@@ -205,7 +205,7 @@ AlphaX 是一个以 `Python + FastAPI + PostgreSQL + Docker + 静态 HTML` 组
 - `app/db/recommendation_queries.py`
   - 推荐热路径查询、active/deduped 查询；不应反向依赖 `altcoin_db.py`。
 - `app/db/push_queries.py`
-  - 推送冷却去重、推送日志、推送前单条推荐读取；推送层只能消费这里派生后的主链路口径。
+  - 推送冷却去重、推送日志、推送前单条推荐读取；推送层只能消费这里派生后的统一状态口径。
 - `app/db/tracking_queries.py`
   - 最新价格缓存、推荐跟踪价格/PnL 写入、入场时点更新。
 - `app/db/cron_queries.py`
@@ -296,7 +296,7 @@ AlphaX 是一个以 `Python + FastAPI + PostgreSQL + Docker + 静态 HTML` 组
 - `/docker`
   - 容器入口与调度器
 - `/tools`
-  - 非主链路工具脚本，如回测和输出摘要脚本
+  - 非运行服务工具脚本，如回测和输出摘要脚本
 - `/templates`
   - 后端读取的 HTML 模板资源
 - `/docs`
@@ -423,7 +423,7 @@ docker compose run --rm alphax-web python scripts/postgres/validate_import.py --
 - schema 变化必须通过 `app/db/migrations/*.sql`。
 - 查询最新运行状态优先看 PostgreSQL 表，而不是历史文件。
 - Docker 容器内运行和宿主机运行可能使用不同连接地址，排查时先确认 `DATABASE_URL`。
-- 调度器并发运行时要检查 lock group，避免多个任务同时写推荐主链路。
+- 调度器并发运行时要检查 lock group，避免多个任务同时写推荐状态机和策略交易账本。
 
 ### 9.3 状态机不要各写各的
 
@@ -448,7 +448,7 @@ docker compose run --rm alphax-web python scripts/postgres/validate_import.py --
 
 ### 9.4 推荐链路当前特别注意点
 
-当前主链路已经能持续产生筛选和确认样本，但后半段仍需要重点盯住：
+当前多策略发现与确认链路已经能持续产生筛选和确认样本，但后半段仍需要重点盯住：
 
 - `latest_price_cache` 可能是实时的，但不代表 `recommendation.pnl_pct` 已更新。
 - `price_tracking` 是跟踪流水，不应和 `latest_price_cache` 混为一谈。
@@ -507,6 +507,11 @@ docker compose run --rm alphax-web python scripts/postgres/validate_import.py --
 4. 修改后至少补 1 个相关测试，最好补到最接近业务口径的那层。
 5. 如果变更影响推荐状态或展示桶，务必同时检查 API、前端、推送、paper trading、历史统计五个面。
 6. 如果变更影响调度任务，务必检查 `scheduler_job_config`、`scheduler_runtime_status` 和最近 `cron_run_log`。
+7. 不要把“底座已建好”当成“功能已完成”。涉及策略、复盘、交易、UI 的任务必须形成端到端闭环：数据写入、状态流转、API、页面展示、操作入口、测试和文档至少各检查一次。
+8. 少问用户开放问题。除非存在资金安全、真实下单、删除数据、不可逆迁移等高风险选择，否则应基于当前代码和产品目标直接推进，并在结果里说明假设和取舍。
+9. 如果发现自己只完成了半成品，不能用“后续可以”收尾；应继续把缺口补完，或明确标记为阻塞并说明为什么当前环境无法完成。
+10. 多策略相关改动必须同时回答三个问题：每个策略是否独立产生信号、是否能独立进入/退出策略交易、是否能独立复盘评价并给出保留/灰度/暂停建议。
+11. UI 层必须与后端能力对等。后端支持多策略筛选、评价或状态时，页面也要提供可理解的筛选、展示和操作入口，不能只返回隐藏字段。
 
 ---
 

app/analysis/reverse_analysis.py

@@ -425,7 +425,7 @@ def discover_new_rules(pattern_summary, all_features, sector_alignments, signifi
         if not rule:
             continue
 
-        # 新体系：逆向分析只生成候选规则，不直接写 learned_rules，避免涨幅榜小样本污染主策略。
+        # 新体系：逆向分析只生成候选规则，不直接写 learned_rules，避免涨幅榜小样本污染已发布策略。
         rule["candidate_id"] = upsert_strategy_rule_candidate(
             source="reverse_analysis",
             rule_type=rule.get("type", "bonus"),

app/config/config_loader.py

@@ -425,7 +425,7 @@ def promote_candidate_rule_to_learned_rule(candidate, release_version=""):
     """把通过发布门槛的候选规则正式写入 learned_rules。
 
     候选规则来自 DB strategy_rule_candidate；只有发布闸门通过时才调用，
-    避免日常研究直接污染主策略。
+    避免日常研究直接污染已发布策略。
     """
     desc = (candidate.get("rule_description") or "").strip()
     if not desc:

app/core/strategy_registry.py

@@ -8,6 +8,9 @@ from dataclasses import dataclass, field
 MAIN_COMPOSITE_STRATEGY = "main_composite_v1"
 BOX_RETEST_1H_STRATEGY = "box_retest_1h_v1"
 BOX_RETEST_4H_STRATEGY = "box_retest_4h_v1"
+VOLUME_IGNITION_1H_STRATEGY = "volume_ignition_1h_v1"
+COMPRESSION_BREAKOUT_4H_STRATEGY = "compression_breakout_4h_v1"
+INTRADAY_MOMENTUM_15M_STRATEGY = "intraday_momentum_15m_v1"
 
 
 @dataclass(frozen=True)
@@ -24,8 +27,8 @@ class StrategyDefinition:
 STRATEGY_DEFINITIONS: dict[str, StrategyDefinition] = {
     MAIN_COMPOSITE_STRATEGY: StrategyDefinition(
         strategy_code=MAIN_COMPOSITE_STRATEGY,
-        strategy_name="综合确认主链路",
+        strategy_name="综合确认策略",
-        description="迁移期兼容主链路，承载现有综合筛选与确认逻辑。",
+        description="迁移期兼容综合策略，承载现有综合筛选与确认逻辑；它与其他策略平等运行。",
         mode="paper_enabled",
     ),
     BOX_RETEST_1H_STRATEGY: StrategyDefinition(
@@ -72,6 +75,70 @@ STRATEGY_DEFINITIONS: dict[str, StrategyDefinition] = {
             "dynamic_leverage_min": 3,
         },
     ),
+    VOLUME_IGNITION_1H_STRATEGY: StrategyDefinition(
+        strategy_code=VOLUME_IGNITION_1H_STRATEGY,
+        strategy_name="1H放量突破启动",
+        description="1H量价齐飞或连续放量后的启动策略，适合捕捉日内到3天的第一段加速。",
+        mode="paper_only",
+        entry_gate_config={
+            "min_entry_score_buy_now": 2,
+            "min_entry_score_wait_pullback": 1,
+            "min_rr_buy_now": 1.25,
+            "breakout_distance_wait_pct": 12,
+            "gain_24h_wait_pct": 10,
+        },
+        paper_config={
+            "entry_min_rr": 1.5,
+            "order_min_rr": 1.5,
+            "order_min_distance_to_entry_pct": 0,
+            "order_require_current_trigger": True,
+            "dynamic_leverage_enabled": True,
+            "dynamic_leverage_min": 3,
+        },
+    ),
+    COMPRESSION_BREAKOUT_4H_STRATEGY: StrategyDefinition(
+        strategy_code=COMPRESSION_BREAKOUT_4H_STRATEGY,
+        strategy_name="4H压缩蓄力突破",
+        description="4H静K蓄力、底部抬高或压缩放量后的突破策略，偏向捕捉1周以内主升前段。",
+        mode="paper_only",
+        entry_gate_config={
+            "min_entry_score_buy_now": 2,
+            "min_entry_score_wait_pullback": 0,
+            "min_rr_buy_now": 1.3,
+            "max_wait_pullback_deviation_pct": 18,
+            "breakout_distance_wait_pct": 18,
+            "gain_24h_wait_pct": 12,
+        },
+        paper_config={
+            "entry_min_rr": 1.6,
+            "order_min_rr": 1.6,
+            "order_min_distance_to_entry_pct": 0,
+            "order_require_current_trigger": False,
+            "dynamic_leverage_enabled": True,
+            "dynamic_leverage_min": 3,
+        },
+    ),
+    INTRADAY_MOMENTUM_15M_STRATEGY: StrategyDefinition(
+        strategy_code=INTRADAY_MOMENTUM_15M_STRATEGY,
+        strategy_name="15m日内动量延续",
+        description="短周期放量突破与1H背景共振的日内动量策略，只做当前触发，不做纯观察追高。",
+        mode="paper_only",
+        entry_gate_config={
+            "min_entry_score_buy_now": 3,
+            "min_entry_score_wait_pullback": 2,
+            "min_rr_buy_now": 1.4,
+            "breakout_distance_wait_pct": 8,
+            "gain_24h_wait_pct": 8,
+        },
+        paper_config={
+            "entry_min_rr": 1.6,
+            "order_min_rr": 1.6,
+            "order_min_distance_to_entry_pct": 0,
+            "order_require_current_trigger": True,
+            "dynamic_leverage_enabled": True,
+            "dynamic_leverage_min": 3,
+        },
+    ),
 }
 
 

app/db/migrations/0015_multi_strategy.sql

@@ -65,7 +65,7 @@ CREATE INDEX IF NOT EXISTS idx_paper_orders_strategy_code
 INSERT INTO strategy_catalog (
     strategy_code, strategy_name, strategy_version, status, mode, description, config_json, created_at, updated_at
 ) VALUES
-    ('main_composite_v1', '综合确认主链路', '', 'active', 'paper_enabled', '迁移期兼容主链路：统一承载旧的综合筛选与确认逻辑。', '{}', NOW()::TEXT, NOW()::TEXT),
+    ('main_composite_v1', '综合确认策略', '', 'active', 'paper_enabled', '迁移期兼容综合策略：统一承载旧的综合筛选与确认逻辑，与其他策略平等运行。', '{}', NOW()::TEXT, NOW()::TEXT),
     ('box_retest_4h_v1', '4H箱体突破回踩', '', 'active', 'paper_only', '底部箱体突破后回踩箱体上沿或EMA承接的结构策略雏形。', '{}', NOW()::TEXT, NOW()::TEXT)
 ON CONFLICT(strategy_code) DO UPDATE SET
     strategy_name=EXCLUDED.strategy_name,

app/db/migrations/0017_altcoin_parallel_strategies.sql

@@ -0,0 +1,12 @@
+INSERT INTO strategy_catalog (
+    strategy_code, strategy_name, strategy_version, status, mode, description, config_json, created_at, updated_at
+) VALUES
+    ('volume_ignition_1h_v1', '1H放量突破启动', '', 'active', 'paper_only', '1H量价齐飞或连续放量后的启动策略，捕捉日内到3天的第一段加速。', '{}', NOW()::TEXT, NOW()::TEXT),
+    ('compression_breakout_4h_v1', '4H压缩蓄力突破', '', 'active', 'paper_only', '4H静K蓄力、底部抬高或压缩放量后的突破策略，偏向捕捉1周以内主升前段。', '{}', NOW()::TEXT, NOW()::TEXT),
+    ('intraday_momentum_15m_v1', '15m日内动量延续', '', 'active', 'paper_only', '短周期放量突破与1H背景共振的日内动量策略，只做当前触发，不做纯观察追高。', '{}', NOW()::TEXT, NOW()::TEXT)
+ON CONFLICT(strategy_code) DO UPDATE SET
+    strategy_name=EXCLUDED.strategy_name,
+    status=EXCLUDED.status,
+    mode=EXCLUDED.mode,
+    description=EXCLUDED.description,
+    updated_at=NOW()::TEXT;

app/db/migrations/0018_strategy_equal_status_labels.sql

@@ -0,0 +1,5 @@
+UPDATE strategy_catalog
+SET strategy_name='综合确认策略',
+    description='迁移期兼容综合策略：统一承载旧的综合筛选与确认逻辑，与其他策略平等运行。',
+    updated_at=NOW()::TEXT
+WHERE strategy_code='main_composite_v1';

app/db/paper_trading.py

@@ -1956,11 +1956,12 @@ def list_paper_orders(limit: int = 50, offset: int = 0, status: str = "", strate
     }
 
 
-def list_paper_trade_events(limit: int = 80, offset: int = 0, symbol: str = "", event_type: str = "") -> dict:
+def list_paper_trade_events(limit: int = 80, offset: int = 0, symbol: str = "", event_type: str = "", strategy_code: str = "") -> dict:
     limit = max(1, min(_safe_int(limit, 80), 200))
     offset = max(0, _safe_int(offset, 0))
     symbol = str(symbol or "").strip().upper()
     event_type = str(event_type or "").strip()
+    strategy_code = str(strategy_code or "").strip()
     where = []
     params = []
     if symbol:
@@ -1969,11 +1970,19 @@ def list_paper_trade_events(limit: int = 80, offset: int = 0, symbol: str = "",
     if event_type:
         where.append("e.event_type=%s")
         params.append(event_type)
+    if strategy_code:
+        where.append("COALESCE(NULLIF(e.strategy_code, ''), t.strategy_code)=%s")
+        params.append(normalize_strategy_code(strategy_code))
     where_sql = "WHERE " + " AND ".join(where) if where else ""
     conn = get_conn()
     try:
         total = conn.execute(
-            f"SELECT COUNT(*) FROM paper_trade_events e {where_sql}",
+            f"""
+            SELECT COUNT(*)
+            FROM paper_trade_events e
+            LEFT JOIN paper_trades t ON t.id = e.trade_id
+            {where_sql}
+            """,
             tuple(params),
         ).fetchone()[0]
         rows = conn.execute(

app/db/recommendation_state.py

@@ -28,7 +28,7 @@ def derive_minimal_state_fields(status, action_status, entry_plan=None):
     action = normalize_action_status(action_status, status)
     if action == "可即刻买入":
         execution_status = "buy_now"
-        reason = "主链路确认当前入场窗口"
+        reason = "策略确认当前入场窗口"
     elif action == "等回踩":
         execution_status = "wait_pullback"
         reason = "等待回踩触发，未触发前不计推荐收益"
@@ -169,7 +169,7 @@ def execution_fields_from_persisted_state(item, entry_plan=None):
         return "invalid", "🔴 已失效，勿追", reason
     if execution_status == "buy_now":
         stop = str(entry_plan.get("stop_loss", "")) if entry_plan else ""
-        return "buy_now", "🟢 现在可买", "推荐时就是可即刻买入；主链路确认当前仍在入场窗口" + (("，风险边界 " + stop) if stop else "")
+        return "buy_now", "🟢 现在可买", "推荐时就是可即刻买入；策略确认当前仍在入场窗口" + (("，风险边界 " + stop) if stop else "")
     if execution_status == "wait_pullback":
         gate = entry_plan.get("entry_quality_gate") or {}
         if gate.get("reasons"):

app/db/review_center.py

@@ -13,7 +13,7 @@ from datetime import datetime, timedelta
 
 from app.db.paper_trading import get_paper_trading_summary
 from app.db.schema import get_conn
-from app.db.strategy_insights import get_strategy_insights
+from app.db.strategy_insights import get_strategy_evaluation, get_strategy_insights
 
 
 def _safe_int(value, default=0):
@@ -387,6 +387,7 @@ def get_review_center_dashboard(days=30):
         paper = _paper_review(conn, since, days)
         evidence = _evidence_review(conn, since)
         iteration = _iteration_review(conn, since)
+        strategy_evaluation = get_strategy_evaluation(days=days)
     finally:
         conn.close()
 
@@ -400,6 +401,7 @@ def get_review_center_dashboard(days=30):
             "策略迭代只发布经过样本约束和灰度闸门验证的规则。",
         ],
         "opportunity": opportunity,
+        "strategy_evaluation": strategy_evaluation,
         "paper_trading": paper,
         "evidence": evidence,
         "iteration": iteration,

app/db/strategy_insights.py

@@ -2,8 +2,9 @@
 
 import json
 import re
+from datetime import datetime, timedelta
 
-from app.core.strategy_registry import normalize_strategy_code, strategy_label
+from app.core.strategy_registry import normalize_strategy_code, registered_strategy_codes, strategy_definition, strategy_label
 from app.db.schema import get_conn
 
 
@@ -31,6 +32,297 @@ def safe_dict_json(value):
     return {}
 
 
+def _safe_float(value, default=0.0):
+    try:
+        if value is None or value == "":
+            return default
+        return float(value)
+    except Exception:
+        return default
+
+
+def _safe_int(value, default=0):
+    try:
+        if value is None or value == "":
+            return default
+        return int(value)
+    except Exception:
+        return default
+
+
+def _pct(part, total):
+    return round(float(part or 0) / float(total or 0) * 100, 2) if total else 0.0
+
+
+def evaluate_strategy_decision(metrics: dict) -> dict:
+    """Turn strategy metrics into an explicit lifecycle recommendation.
+
+    This is advisory only. It does not mutate strategy configs; release/pause
+    should still go through the strategy iteration gate.
+    """
+    signal_count = _safe_int(metrics.get("signal_count"))
+    opportunity_count = _safe_int(metrics.get("opportunity_count"))
+    trade_count = _safe_int(metrics.get("trade_count"))
+    closed_count = _safe_int(metrics.get("closed_trade_count"))
+    win_rate = _safe_float(metrics.get("win_rate_pct"))
+    avg_pnl = _safe_float(metrics.get("avg_realized_pnl_pct"))
+    realized = _safe_float(metrics.get("realized_pnl_usdt"))
+    worst = _safe_float(metrics.get("worst_pnl_pct"))
+    fill_rate = _safe_float(metrics.get("order_fill_rate_pct"))
+    trade_conversion = _safe_float(metrics.get("trade_conversion_pct"))
+
+    score = 50.0
+    score += min(signal_count, 40) * 0.25
+    score += min(opportunity_count, 40) * 0.2
+    score += min(trade_conversion, 40) * 0.25
+    score += (win_rate - 50) * 0.35 if closed_count else 0
+    score += max(-12, min(12, avg_pnl)) * 1.8
+    score += 6 if realized > 0 else (-6 if realized < 0 else 0)
+    score += 4 if fill_rate >= 40 else 0
+    score -= 8 if worst <= -8 else 0
+    score = round(max(0, min(100, score)), 1)
+
+    reasons = []
+    next_actions = []
+    decision = "observe"
+    decision_label = "继续观察"
+
+    if signal_count < 5 and opportunity_count < 5:
+        decision = "collect_samples"
+        decision_label = "样本不足"
+        reasons.append("信号和机会样本不足，暂不判断优劣")
+        next_actions.append("继续收集样本，不要直接调高权重")
+    elif trade_count == 0:
+        decision = "review_entry_gate"
+        decision_label = "检查入场闸门"
+        reasons.append("已有发现样本，但还没有进入策略交易")
+        next_actions.append("检查 RR、买点距离、全局风控和挂单成交条件")
+    elif closed_count < 5:
+        decision = "gray"
+        decision_label = "灰度观察"
+        reasons.append("已有交易样本，但平仓数量不足 5 笔")
+        next_actions.append("保持 paper-only，等更多已平仓样本")
+    elif win_rate >= 55 and avg_pnl > 0 and realized > 0:
+        decision = "promote"
+        decision_label = "优先保留"
+        reasons.append("胜率、平均收益和已实现收益同时为正")
+        next_actions.append("允许维持或小幅提升策略权重，但仍需观察回撤")
+    elif win_rate < 35 or avg_pnl <= -2 or realized < 0 and worst <= -6:
+        decision = "pause"
+        decision_label = "暂停/降权"
+        reasons.append("胜率或平均收益不达标，且存在较差回撤")
+        next_actions.append("暂停新增真实跟单，只保留观察或降低仓位")
+    elif fill_rate < 15 and opportunity_count >= 10:
+        decision = "tune_entry"
+        decision_label = "优化入场"
+        reasons.append("机会样本不少，但挂单成交或执行转化偏低")
+        next_actions.append("复查挂单价格、有效期、回踩距离和成交触发")
+    else:
+        decision = "keep"
+        decision_label = "保留运行"
+        reasons.append("当前表现没有触发暂停或升级条件")
+        next_actions.append("继续按当前门槛运行并积累样本")
+
+    return {
+        "decision": decision,
+        "decision_label": decision_label,
+        "evaluation_score": score,
+        "reasons": reasons,
+        "next_actions": next_actions,
+    }
+
+
+def get_strategy_evaluation(days: int = 30) -> dict:
+    days = max(1, min(_safe_int(days, 30), 365))
+    since = (datetime.now() - timedelta(days=days)).isoformat()
+    codes = registered_strategy_codes()
+    metrics = {}
+    for code in codes:
+        definition = strategy_definition(code)
+        metrics[code] = {
+            "strategy_code": code,
+            "strategy_name": definition.strategy_name,
+            "description": definition.description,
+            "mode": definition.mode,
+            "status": definition.status,
+            "signal_count": 0,
+            "candidate_signal_count": 0,
+            "observe_signal_count": 0,
+            "avg_signal_confidence": 0.0,
+            "opportunity_count": 0,
+            "actionable_count": 0,
+            "buy_now_count": 0,
+            "wait_pullback_count": 0,
+            "observe_count": 0,
+            "order_count": 0,
+            "filled_order_count": 0,
+            "canceled_order_count": 0,
+            "trade_count": 0,
+            "open_trade_count": 0,
+            "closed_trade_count": 0,
+            "win_count": 0,
+            "loss_count": 0,
+            "realized_pnl_usdt": 0.0,
+            "pnl_pct_values": [],
+            "best_pnl_pct": None,
+            "worst_pnl_pct": None,
+        }
+
+    def bucket(code):
+        normalized = normalize_strategy_code(code)
+        if normalized not in metrics:
+            definition = strategy_definition(normalized)
+            metrics[normalized] = {
+                "strategy_code": normalized,
+                "strategy_name": definition.strategy_name,
+                "description": definition.description,
+                "mode": definition.mode,
+                "status": definition.status,
+                "signal_count": 0,
+                "candidate_signal_count": 0,
+                "observe_signal_count": 0,
+                "avg_signal_confidence": 0.0,
+                "opportunity_count": 0,
+                "actionable_count": 0,
+                "buy_now_count": 0,
+                "wait_pullback_count": 0,
+                "observe_count": 0,
+                "order_count": 0,
+                "filled_order_count": 0,
+                "canceled_order_count": 0,
+                "trade_count": 0,
+                "open_trade_count": 0,
+                "closed_trade_count": 0,
+                "win_count": 0,
+                "loss_count": 0,
+                "realized_pnl_usdt": 0.0,
+                "pnl_pct_values": [],
+                "best_pnl_pct": None,
+                "worst_pnl_pct": None,
+            }
+        return metrics[normalized]
+
+    conn = get_conn()
+    try:
+        for row in conn.execute(
+            """
+            SELECT strategy_code,
+                   COUNT(*) AS signal_count,
+                   SUM(CASE WHEN signal_status='candidate' THEN 1 ELSE 0 END) AS candidate_count,
+                   SUM(CASE WHEN signal_status='observe' THEN 1 ELSE 0 END) AS observe_count,
+                   AVG(confidence) AS avg_confidence
+            FROM strategy_signals
+            WHERE created_at >= %s
+            GROUP BY strategy_code
+            """,
+            (since,),
+        ).fetchall():
+            b = bucket(row.get("strategy_code"))
+            b["signal_count"] = _safe_int(row.get("signal_count"))
+            b["candidate_signal_count"] = _safe_int(row.get("candidate_count"))
+            b["observe_signal_count"] = _safe_int(row.get("observe_count"))
+            b["avg_signal_confidence"] = round(_safe_float(row.get("avg_confidence")), 2)
+
+        for row in conn.execute(
+            """
+            SELECT strategy_code,
+                   COUNT(*) AS opportunity_count,
+                   SUM(CASE WHEN execution_status IN ('buy_now','wait_pullback') THEN 1 ELSE 0 END) AS actionable_count,
+                   SUM(CASE WHEN execution_status='buy_now' THEN 1 ELSE 0 END) AS buy_now_count,
+                   SUM(CASE WHEN execution_status='wait_pullback' THEN 1 ELSE 0 END) AS wait_pullback_count,
+                   SUM(CASE WHEN execution_status='observe' THEN 1 ELSE 0 END) AS observe_count
+            FROM recommendation
+            WHERE rec_time >= %s
+            GROUP BY strategy_code
+            """,
+            (since,),
+        ).fetchall():
+            b = bucket(row.get("strategy_code"))
+            for key in ("opportunity_count", "actionable_count", "buy_now_count", "wait_pullback_count", "observe_count"):
+                b[key] = _safe_int(row.get(key))
+
+        for row in conn.execute(
+            """
+            SELECT strategy_code,
+                   COUNT(*) AS order_count,
+                   SUM(CASE WHEN status='filled' THEN 1 ELSE 0 END) AS filled_count,
+                   SUM(CASE WHEN status IN ('canceled','rejected','expired') THEN 1 ELSE 0 END) AS canceled_count
+            FROM paper_orders
+            WHERE created_at >= %s
+            GROUP BY strategy_code
+            """,
+            (since,),
+        ).fetchall():
+            b = bucket(row.get("strategy_code"))
+            b["order_count"] = _safe_int(row.get("order_count"))
+            b["filled_order_count"] = _safe_int(row.get("filled_count"))
+            b["canceled_order_count"] = _safe_int(row.get("canceled_count"))
+
+        for row in conn.execute(
+            """
+            SELECT strategy_code, status, realized_pnl_pct, realized_pnl_usdt, pnl_pct
+            FROM paper_trades
+            WHERE opened_at >= %s
+            """,
+            (since,),
+        ).fetchall():
+            b = bucket(row.get("strategy_code"))
+            status = row.get("status") or ""
+            b["trade_count"] += 1
+            if status == "open":
+                b["open_trade_count"] += 1
+            if status == "closed":
+                b["closed_trade_count"] += 1
+                pnl_pct = _safe_float(row.get("realized_pnl_pct"))
+                pnl_usdt = _safe_float(row.get("realized_pnl_usdt"))
+                b["realized_pnl_usdt"] += pnl_usdt
+                b["pnl_pct_values"].append(pnl_pct)
+                if pnl_pct > 0:
+                    b["win_count"] += 1
+                elif pnl_pct < 0:
+                    b["loss_count"] += 1
+                b["best_pnl_pct"] = pnl_pct if b["best_pnl_pct"] is None else max(b["best_pnl_pct"], pnl_pct)
+                b["worst_pnl_pct"] = pnl_pct if b["worst_pnl_pct"] is None else min(b["worst_pnl_pct"], pnl_pct)
+    finally:
+        conn.close()
+
+    strategies = []
+    for item in metrics.values():
+        values = item.pop("pnl_pct_values", [])
+        item["realized_pnl_usdt"] = round(item["realized_pnl_usdt"], 4)
+        item["avg_realized_pnl_pct"] = round(sum(values) / len(values), 4) if values else 0.0
+        item["win_rate_pct"] = _pct(item["win_count"], item["closed_trade_count"])
+        item["actionable_rate_pct"] = _pct(item["actionable_count"], item["opportunity_count"])
+        item["trade_conversion_pct"] = _pct(item["trade_count"], item["opportunity_count"])
+        item["order_fill_rate_pct"] = _pct(item["filled_order_count"], item["order_count"])
+        item["candidate_signal_rate_pct"] = _pct(item["candidate_signal_count"], item["signal_count"])
+        item["best_pnl_pct"] = item["best_pnl_pct"] if item["best_pnl_pct"] is not None else 0.0
+        item["worst_pnl_pct"] = item["worst_pnl_pct"] if item["worst_pnl_pct"] is not None else 0.0
+        item.update(evaluate_strategy_decision(item))
+        strategies.append(item)
+
+    strategies.sort(key=lambda x: (x["evaluation_score"], x["realized_pnl_usdt"], x["closed_trade_count"], x["signal_count"]), reverse=True)
+    decisions = {}
+    for item in strategies:
+        decisions[item["decision"]] = decisions.get(item["decision"], 0) + 1
+    return {
+        "definition": "策略评价按 strategy_code 独立统计发现、执行、成交、收益和回撤，并给出保留/灰度/暂停等建议；建议不直接改配置，仍需经过策略迭代闸门。",
+        "days": days,
+        "generated_at": datetime.now().isoformat(timespec="seconds"),
+        "summary": {
+            "strategy_count": len(strategies),
+            "active_signal_strategy_count": sum(1 for x in strategies if x["signal_count"] > 0),
+            "traded_strategy_count": sum(1 for x in strategies if x["trade_count"] > 0),
+            "promote_count": decisions.get("promote", 0),
+            "pause_count": decisions.get("pause", 0),
+            "gray_count": decisions.get("gray", 0),
+            "collect_samples_count": decisions.get("collect_samples", 0),
+        },
+        "decision_distribution": [{"name": k, "count": v} for k, v in sorted(decisions.items(), key=lambda x: (-x[1], x[0]))],
+        "strategies": strategies,
+    }
+
+
 def get_strategy_insights():
     """Strategy attribution based on opportunity and paper-trading conversion.
 

app/services/altcoin_confirm.py

@@ -54,6 +54,11 @@ from app.core.market_regime import classify_market_regime
 from app.db.onchain_db import get_onchain_factor_context
 from app.db.strategy_signal_queries import insert_strategy_signal
 from app.services.market_overview import get_crypto_market_overview
+from app.strategies.altcoin_breakout import (
+    build_compression_breakout_4h_signal,
+    build_intraday_momentum_15m_signal,
+    build_volume_ignition_1h_signal,
+)
 from app.strategies.box_retest_4h import build_box_retest_1h_signal, build_box_retest_4h_signal
 from app.config.config_loader import _get_section as _get_cfg_section
 from app.core.pa_engine import (
@@ -85,10 +90,13 @@ def _strategy_context_for_recommendation(symbol: str, result: dict, entry_plan:
     """Build and persist a standard strategy signal when an independent strategy matches."""
     bp_1h = result.get("box_breakout_pullback_1h") or (result.get("market_context") or {}).get("box_breakout_pullback_1h") or {}
     bp_4h = result.get("box_breakout_pullback_4h") or (result.get("market_context") or {}).get("box_breakout_pullback_4h") or {}
-    if not bp_1h.get("detected") and not bp_4h.get("detected"):
-        return {}
     market_regime = result.get("market_regime") or (result.get("market_context") or {}).get("market_regime") or {}
     signal_candidates = []
+    signal_candidates.extend([
+        build_volume_ignition_1h_signal(symbol=symbol, result=result, entry_plan=entry_plan or {}),
+        build_compression_breakout_4h_signal(symbol=symbol, result=result, entry_plan=entry_plan or {}),
+        build_intraday_momentum_15m_signal(symbol=symbol, result=result, entry_plan=entry_plan or {}),
+    ])
     if bp_1h.get("detected"):
         signal_candidates.append(
             build_box_retest_1h_signal(
@@ -2074,8 +2082,8 @@ def main(compact: bool = False, verbose: bool = False, limit: int | None = None,
                     detail={**cand_detail, **result},
                 )
                 result["state_update"] = state_result
-                # 飞书只是通知层：确认阶段不再绕过 recommendation 主链路直接推送。
+                # 飞书只是通知层：确认阶段不再绕过 recommendation 状态机直接推送。
-                # 先完成 create_recommendation + DB 主状态派生，再用同一条主链路结果决定是否通知。
+                # 先完成 create_recommendation + DB 状态派生，再用同一条状态结果决定是否通知。
 
                 # 🟢 只做做多！方向永远多头
                 rec_direction = get_strategy_direction()

app/services/chat_assistant.py

@@ -504,7 +504,7 @@ def _fallback_answer(intent: str, message: str, context: dict) -> dict:
         evidence.append(f"技术面：{tech_summary.get('headline', '')}")
     if recommendations:
         r = recommendations[0]
-        evidence.append(f"主链路：{r.get('execution_label') or r.get('action_status') or r.get('execution_status')}，原因：{r.get('execution_reason') or r.get('state_reason') or '--'}")
+        evidence.append(f"策略状态：{r.get('execution_label') or r.get('action_status') or r.get('execution_status')}，原因：{r.get('execution_reason') or r.get('state_reason') or '--'}")
     if sentiment:
         evidence.append(f"舆情：近 72h 有 {len(sentiment)} 条相关事件，最新为「{sentiment[0].get('title', '')[:60]}」。")
     if isinstance(onchain, dict) and (onchain.get("events") or onchain.get("metrics")):

app/services/price_tracker.py

@@ -321,8 +321,8 @@ def track_prices():
             # PA增强：动态跟踪信号分析
             tracking_signals = analyze_tracking_signals(symbol, rec, current_price)
 
-            # 主链路状态迁移：tracker 只提交“候选状态 + 当前价”，最终状态由 DB 主链路统一落库。
+            # 统一状态迁移：tracker 只提交“候选状态 + 当前价”，最终状态由 DB 状态机统一落库。
-            # 飞书推送只能消费主链路返回的最终状态，不能再自行判断。
+            # 飞书推送只能消费统一状态机返回的最终状态，不能再自行判断。
             terminal_action = {
                 "hit_tp2": "止盈2",
                 "stopped_out": "止损",

app/services/review_engine.py

@@ -908,7 +908,7 @@ def _extract_rules_from_review():
                 "score_adjust": 2 if len(combo) >= 2 else 1,
                 "source": "review_pattern",
             }
-            # 新体系：先进入候选规则池，不直接污染主策略。达到发布门槛后再升级为 active。
+            # 新体系：先进入候选规则池，不直接污染已发布策略。达到发布门槛后再升级为 active。
             rule["candidate_id"] = upsert_strategy_rule_candidate(
                 source="review_pattern",
                 rule_type=rule.get("type", "bonus"),
@@ -1215,7 +1215,7 @@ def _build_dual_attribution(results, effect_summary):
             confidence_score=round(min(90, 45 + cnt * 10), 1),
             sample_size=cnt,
             status="candidate",
-            notes="失败归因生成：先入候选池，不立即改主策略",
+            notes="失败归因生成：先入候选池，不立即改已发布策略",
             source_ref=f"failure:{ftype}",
         )
         candidate_rules.append({"id": cid, "type": "penalty", "signal": ftype, "description": desc, "confidence_score": round(min(90, 45 + cnt * 10), 1), "sample_size": cnt, "status": "candidate"})

app/strategies/altcoin_breakout.py

@@ -0,0 +1,176 @@
+"""Altcoin breakout strategy profiles.
+
+These builders turn existing evidence into complete strategy signals. A factor
+can support a strategy, but the strategy still owns trigger freshness, entry
+quality and risk semantics.
+"""
+
+from __future__ import annotations
+
+from app.core.factor_roles import CONFIRMATION, ENTRY, PREREQUISITE, RISK, TRIGGER
+from app.core.strategy_contract import StrategySignal, current_strategy_version
+from app.core.strategy_registry import (
+    COMPRESSION_BREAKOUT_4H_STRATEGY,
+    INTRADAY_MOMENTUM_15M_STRATEGY,
+    VOLUME_IGNITION_1H_STRATEGY,
+)
+
+
+def _safe_float(value, default=0.0) -> float:
+    try:
+        if value is None or value == "":
+            return default
+        return float(value)
+    except Exception:
+        return default
+
+
+def _signals_text(result: dict) -> str:
+    return " ".join(str(x or "") for x in (result or {}).get("signals") or [])
+
+
+def _trigger_context(result: dict) -> dict:
+    return (result or {}).get("trigger_context") or ((result or {}).get("market_context") or {}).get("trigger_context") or {}
+
+
+def _has_current_trigger(result: dict) -> bool:
+    trigger = _trigger_context(result)
+    text = _signals_text(result)
+    return bool(trigger.get("current_triggers")) or "15min即刻入场" in text or "15min强突破" in text or "15min 强突破" in text
+
+
+def _status_for_entry(result: dict, entry_plan: dict | None = None, *, require_current_trigger: bool = False, allow_wait: bool = True) -> tuple[str, list[str]]:
+    reasons = []
+    entry_action = str((entry_plan or {}).get("entry_action") or ((result or {}).get("entry_plan") or {}).get("entry_action") or (result or {}).get("entry_action") or "").strip()
+    if require_current_trigger and not _has_current_trigger(result):
+        return "observe", ["缺少当前低周期触发"]
+    if entry_action in ("可即刻买入", "即刻买入"):
+        return "candidate", reasons
+    if entry_action == "等回踩" and allow_wait:
+        return "candidate", reasons
+    if entry_action:
+        reasons.append(f"当前入场动作 {entry_action} 不满足策略执行条件")
+    return "observe", reasons
+
+
+def build_volume_ignition_1h_signal(*, symbol: str, result: dict, entry_plan: dict | None = None) -> StrategySignal | None:
+    text = _signals_text(result)
+    has_vp = "量价齐飞" in text or ("连续" in text and "放量" in text)
+    has_breakout = "1H" in text and ("突破" in text or "起爆" in text)
+    if not (has_vp or has_breakout):
+        return None
+    status, reasons = _status_for_entry(result, entry_plan, require_current_trigger=False, allow_wait=True)
+    score = _safe_float(result.get("score"))
+    confidence = min(100.0, max(0.0, score * 7 + (12 if _has_current_trigger(result) else 0)))
+    trigger = {
+        "factor_code": "vp_fly_1h_current" if has_vp else "ignition_1h_current",
+        "factor_label": "1H放量突破启动",
+        "has_current_trigger": _has_current_trigger(result),
+        "trigger_status": _trigger_context(result).get("trigger_status") or "",
+        "entry_action": (entry_plan or {}).get("entry_action") or "",
+    }
+    return StrategySignal(
+        strategy_code=VOLUME_IGNITION_1H_STRATEGY,
+        strategy_version=current_strategy_version(),
+        symbol=symbol,
+        direction="long",
+        status=status,
+        confidence=confidence,
+        score=score,
+        trigger=trigger,
+        factor_roles={
+            "vp_fly_1h_current": TRIGGER,
+            "volume_consecutive_1h": CONFIRMATION,
+            "breakout_15m_current": ENTRY,
+            "pullback_15m_confirm": ENTRY,
+            "false_breakout": RISK,
+            "risk_reward_bad": RISK,
+        },
+        entry_plan=entry_plan or {},
+        risk_plan={
+            "invalid_if": ["放量后不能延续", "15m假突破", "跌回启动K低点", "RR不足"],
+            "risk_reasons": reasons,
+        },
+        decision_log={"module": VOLUME_IGNITION_1H_STRATEGY, "decision": status, "reasons": reasons},
+    )
+
+
+def build_compression_breakout_4h_signal(*, symbol: str, result: dict, entry_plan: dict | None = None) -> StrategySignal | None:
+    text = _signals_text(result)
+    has_compression = any(key in text for key in ("静K", "压缩", "布林收窄", "底部抬高"))
+    has_breakout_context = any(key in text for key in ("突破", "起爆", "量价齐飞", "回踩"))
+    if not (has_compression and has_breakout_context):
+        return None
+    status, reasons = _status_for_entry(result, entry_plan, require_current_trigger=False, allow_wait=True)
+    score = _safe_float(result.get("score"))
+    confidence = min(100.0, max(0.0, score * 6 + (10 if "底部" in text else 0) + (8 if _has_current_trigger(result) else 0)))
+    return StrategySignal(
+        strategy_code=COMPRESSION_BREAKOUT_4H_STRATEGY,
+        strategy_version=current_strategy_version(),
+        symbol=symbol,
+        direction="long",
+        status=status,
+        confidence=confidence,
+        score=score,
+        trigger={
+            "factor_code": "compression_surge_4h",
+            "factor_label": "4H压缩蓄力突破",
+            "has_current_trigger": _has_current_trigger(result),
+            "trigger_status": _trigger_context(result).get("trigger_status") or "",
+        },
+        factor_roles={
+            "static_accum_4h": PREREQUISITE,
+            "higher_lows_4h": PREREQUISITE,
+            "compression_surge_4h": TRIGGER,
+            "ignition_4h_current": CONFIRMATION,
+            "vp_fly_1h_current": CONFIRMATION,
+            "pullback_15m_confirm": ENTRY,
+            "false_breakout": RISK,
+        },
+        entry_plan=entry_plan or {},
+        risk_plan={
+            "invalid_if": ["突破后跌回压缩区间", "回踩放量跌破", "无量反抽失败", "市场风险升高"],
+            "risk_reasons": reasons,
+        },
+        decision_log={"module": COMPRESSION_BREAKOUT_4H_STRATEGY, "decision": status, "reasons": reasons},
+    )
+
+
+def build_intraday_momentum_15m_signal(*, symbol: str, result: dict, entry_plan: dict | None = None) -> StrategySignal | None:
+    text = _signals_text(result)
+    has_short_tf = any(key in text for key in ("15min", "15m", "短周期", "5m/15m"))
+    if not has_short_tf:
+        return None
+    status, reasons = _status_for_entry(result, entry_plan, require_current_trigger=True, allow_wait=False)
+    score = _safe_float(result.get("score"))
+    confidence = min(100.0, max(0.0, score * 6 + 18))
+    return StrategySignal(
+        strategy_code=INTRADAY_MOMENTUM_15M_STRATEGY,
+        strategy_version=current_strategy_version(),
+        symbol=symbol,
+        direction="long",
+        status=status,
+        confidence=confidence,
+        score=score,
+        trigger={
+            "factor_code": "short_tf_15m_ignition",
+            "factor_label": "15m日内动量延续",
+            "has_current_trigger": _has_current_trigger(result),
+            "trigger_status": _trigger_context(result).get("trigger_status") or "",
+        },
+        factor_roles={
+            "short_tf_5m_ignition": PREREQUISITE,
+            "short_tf_15m_ignition": TRIGGER,
+            "short_tf_resonance": CONFIRMATION,
+            "vp_fly_1h_current": CONFIRMATION,
+            "breakout_15m_current": ENTRY,
+            "false_breakout": RISK,
+            "trend_exhaustion": RISK,
+        },
+        entry_plan=entry_plan or {},
+        risk_plan={
+            "invalid_if": ["15m跌回突破K", "短周期量能衰减", "快速冲高回落", "RR不足"],
+            "risk_reasons": reasons,
+        },
+        decision_log={"module": INTRADAY_MOMENTUM_15M_STRATEGY, "decision": status, "reasons": reasons},
+    )

app/web/routes_paper_trading.py

@@ -11,6 +11,7 @@ from app.db.paper_trading import (
     reset_paper_trading_data,
     send_paper_trading_report,
 )
+from app.db.strategy_insights import get_strategy_evaluation
 from app.web.shared import require_admin
 
 
@@ -29,6 +30,12 @@ async def api_paper_trading_performance(days: int = 30, altcoin_session: str = C
     return get_paper_trading_performance(days=days)
 
 
+@router.get("/api/paper-trading/strategies")
+async def api_paper_trading_strategies(days: int = 30, altcoin_session: str = Cookie(default="")):
+    require_admin(altcoin_session)
+    return get_strategy_evaluation(days=days)
+
+
 @router.get("/api/paper-trading/trades")
 async def api_paper_trading_trades(
     limit: int = 50,
@@ -59,10 +66,11 @@ async def api_paper_trading_events(
     offset: int = 0,
     symbol: str = "",
     event_type: str = "",
+    strategy_code: str = "",
     altcoin_session: str = Cookie(default=""),
 ):
     require_admin(altcoin_session)
-    return list_paper_trade_events(limit=limit, offset=offset, symbol=symbol, event_type=event_type)
+    return list_paper_trade_events(limit=limit, offset=offset, symbol=symbol, event_type=event_type, strategy_code=strategy_code)
 
 
 @router.post("/api/paper-trading/report")

app/web/routes_strategy.py

@@ -12,6 +12,7 @@ from app.db.review_queries import (
     get_strategy_rule_candidates,
     refresh_strategy_candidate_performance,
 )
+from app.db.strategy_insights import get_strategy_evaluation
 from app.services.llm_insights import get_latest_review_memo
 from app.db.schema import get_conn
 from app.db.altcoin_db import _derive_execution_fields
@@ -72,6 +73,12 @@ async def api_strategy_insights(altcoin_session: str = Cookie(default="")):
     return get_strategy_insights()
 
 
+@router.get("/api/strategy/evaluation")
+async def api_strategy_evaluation(days: int = 30, altcoin_session: str = Cookie(default="")):
+    require_api_user_with_subscription(altcoin_session)
+    return get_strategy_evaluation(days=days)
+
+
 @router.get("/api/strategy/lifecycle")
 async def api_strategy_lifecycle(days: int = 30, altcoin_session: str = Cookie(default="")):
     require_api_user_with_subscription(altcoin_session)

docs/MULTI_STRATEGY_ARCHITECTURE.md

@@ -6,7 +6,7 @@
 
 ## 1. 为什么要改
 
-当前系统已经有 `strategy_version`、`FactorScorer`、`factor_score_breakdown` 和 paper trading 归因，但主链路仍更像一个“大而全评分器”：
+当前系统已经有 `strategy_version`、`FactorScorer`、`factor_score_breakdown` 和 paper trading 归因，但综合确认策略仍更像一个“大而全评分器”：
 
 - 很多因子在同一个确认函数里叠加，容易把单根行情重复加分。
 - 推荐和 paper trading 更容易知道“综合分高不高”，但不容易知道“到底是哪套策略赚了钱”。
@@ -23,7 +23,7 @@
 - 策略输出必须走统一契约，不能每个策略自定义一套不可比的 JSON。
 - 推荐、挂单、持仓、事件日志、复盘都必须保留策略血缘。
 - 因子角色必须显式声明，未知因子不能默认成为交易触发。
-- 旧主链路可以兼容为 `main_composite_v1`，但新数据不允许没有策略来源。
+- 历史综合确认策略可以兼容为 `main_composite_v1`，但它与其他策略平等运行；新数据不允许没有策略来源。
 - 策略中文名、描述、启用状态要集中维护，不能散落在页面和业务代码里。
 - 多策略架构上线后，收益评价必须按策略拆分；总收益只能作为账户层结果，不能替代策略评价。
 
@@ -179,12 +179,12 @@
 - 建立 `app/core/strategy_contract.py`，定义策略输出结构。
 - 建立 `app/core/factor_roles.py`，统一因子角色分类。
 - 给 `recommendation` / `paper_trades` / `paper_orders` 补 `strategy_code` 和 `strategy_signal_id`。
-- 在确认层先把现有主链路标为 `main_composite_v1`。
+- 在确认层先把现有综合确认策略标为 `main_composite_v1`。
 - 把 `box_breakout_pullback_4h` 标记为 `box_retest_4h_v1` 的核心触发候选，但仍通过完整策略条件判断。
 
 ### P1：拆出第一个独立策略
 
-目标：让 `box_retest_4h_v1` 独立运行，与原主链路并行。
+目标：让 `box_retest_4h_v1` 独立运行，与其他策略平等并行。
 
 - 新增 `app/strategies/box_retest_4h.py`。
 - 让它消费统一交易宇宙和 4H K线。

docs/MULTI_STRATEGY_IMPLEMENTATION_PLAN.md

@@ -5,7 +5,7 @@
 今晚目标不是一次性完成所有长期架构，而是交付一个能真实运行的第一阶段：
 
 - 每条推荐、挂单、持仓都能追溯到策略。
-- 现有主链路不被打断，统一标记为 `main_composite_v1`。
+- 现有综合确认策略不被打断，统一标记为 `main_composite_v1`。
 - `box_retest_4h_v1` 进入独立策略雏形，不再只是综合确认层里的一个因子。
 - paper trading 和复盘能按策略看表现。
 - 前端至少能看到策略来源，明早可以观察跑出来的数据。
@@ -27,7 +27,7 @@
 
 允许的兼容：
 
-- 旧主链路继续运行，但必须标记为 `main_composite_v1`。
+- 旧综合确认策略继续运行，但必须标记为 `main_composite_v1`，并与其他策略平等。
 - 第一版 `box_retest_4h_v1` 可以先 paper-only / observe-only。
 - 复杂仲裁可以先做最小规则，但接口要为后续多策略扩展留好。
 
@@ -219,7 +219,7 @@ risk_reward_bad -> risk
 entry_quality_gate -> risk
 ```
 
-注意：同一因子在不同策略里可以有不同角色，但默认映射用于主链路兼容。
+注意：同一因子在不同策略里可以有不同角色；默认映射只用于兼容综合确认策略。
 
 该模块必须保持纯函数，不依赖 DB，不读运行时配置，避免策略基础设施反向依赖业务层。
 
@@ -252,7 +252,7 @@ decision_log: dict
 
 - 提供 `to_json_dict()`，统一序列化。
 - 提供 `from_confirm_result()`，兼容现有确认链路。
-- 提供 `default_main_composite_signal()`，给旧主链路生成标准策略快照。
+- 提供 `default_main_composite_signal()`，给旧综合确认策略生成标准策略快照。
 - 不在这里调用数据库，不在这里拉行情。
 
 ### 3.2.1 新增 `app/db/strategy_signal_queries.py`
@@ -337,7 +337,7 @@ app/core/strategy_registry.py
 初始策略：
 
 ```text
-main_composite_v1 -> 综合确认主链路
+main_composite_v1 -> 综合确认策略
 box_retest_4h_v1 -> 4H箱体突破回踩
 ```
 
@@ -403,15 +403,15 @@ app/strategies/orchestrator.py
 
 - `StrategySignal(strategy_code='box_retest_4h_v1')`
 
-### 4.4 与现有主链路关系
+### 4.4 与现有综合确认策略关系
 
-第一阶段不能直接删除现有主链路里的 `box_breakout_pullback_4h` 加分，否则会改变当前推荐行为过大。
+第一阶段不能直接删除现有综合确认策略里的 `box_breakout_pullback_4h` 加分，否则会改变当前推荐行为过大。
 
 处理方式：
 
-- 主链路继续把它作为结构因子参与确认，策略来源仍为 `main_composite_v1`。
+- 综合确认策略继续把它作为结构因子参与确认，策略来源仍为 `main_composite_v1`。
 - 独立策略并行生成 `box_retest_4h_v1` 信号。
-- 如果同一 symbol 同时被主链路和箱体策略命中，短期由仲裁器保留更明确的策略快照，或在推荐 `strategy_snapshot_json` 中记录 secondary strategies。
+- 如果同一 symbol 同时被综合确认策略和箱体策略命中，短期由仲裁器保留更明确的策略快照，或在推荐 `strategy_snapshot_json` 中记录 secondary strategies。
 - 明早先观察是否出现重复推荐，再决定是否让箱体策略接管该类信号。
 
 ## 5. API 与 UI 改造
@@ -552,11 +552,11 @@ docker compose exec alphax-web python -m app.cli paper-trader
 - 不做复杂策略投票。
 - 不做多空完整重构。
 - 不做大规模 UI 重设计。
-- 不删除现有主链路。
+- 不删除现有综合确认策略。
 
 今晚可以做：
 
-- 保留主链路运行。
+- 保留综合确认策略运行。
 - 补策略血缘。
 - 补第一版策略标准结构。
 - 让 `box_retest_4h_v1` 作为独立候选跑起来。
@@ -568,7 +568,7 @@ docker compose exec alphax-web python -m app.cli paper-trader
 
 1. 新增 migration 和回填。
 2. 新增 `strategy_registry`、`factor_roles`、`strategy_contract` 纯核心模块。
-3. 新增 `strategy_signal_queries`，不要碰业务主链路。
+3. 新增 `strategy_signal_queries`，不要碰业务状态机。
 4. 修改 `create_recommendation()` 写入策略血缘。
 5. 修改 paper orders / paper trades 继承策略血缘。
 6. 修改读模型/API，让前端能看到策略来源。

docs/OPTIMIZATION_TODO.md

@@ -47,7 +47,7 @@
 - 建立 `app/core/strategy_registry.py`，集中维护策略代码、中文名、描述、默认模式，避免策略名散落硬编码。
 - 建立 `app/db/strategy_signal_queries.py` 和 `strategy_signals` 表，保存标准策略信号。
 - 给 `recommendation`、`paper_trades`、`paper_orders` 增加 `strategy_code`、`strategy_signal_id`、`strategy_snapshot_json`、`factor_roles_json` 等策略血缘字段。
-- 现有综合链路先标记为 `main_composite_v1`，避免无策略来源的推荐继续进入账本。
+- 现有综合确认策略先标记为 `main_composite_v1`，它与其他策略平等，避免无策略来源的推荐继续进入账本。
 - 先把 `box_retest_4h_v1` 作为第一个独立策略候选拆出来：`box_breakout_pullback_4h` 只能是核心触发因子，仍要经过市场环境、交易宇宙、确认、入场、风控和失效条件。
 - 复盘中心增加按 `strategy_code` 聚合的胜率、收益、最大回撤、盈亏比和持仓时长。
 - 新数据不得出现空 `strategy_code`；旧数据通过 migration 回填为 `main_composite_v1`，但不能继续产生无来源样本。
@@ -163,7 +163,7 @@
 ## 下一步执行建议
 
 1. 先完成多策略第一阶段的策略血缘字段和标准策略输出结构。
-2. 把现有主链路标记为 `main_composite_v1`，确保 recommendation 和 paper trading 不再丢失策略来源。
+2. 把现有综合确认策略标记为 `main_composite_v1`，确保 recommendation 和 paper trading 不再丢失策略来源。
 3. 拆出 `box_retest_4h_v1` 作为第一个独立策略候选，先 observe/paper-only 跑样本。
 4. 部署运行一段时间，观察 `market_regime`、`score_components`、`factor_score_breakdown.groups`、观察/挂单推进率是否能解释真实回撤。
 5. 按线上样本校准 `entry_score` 门槛、group cap 和 high-risk 门槛。

docs/PROJECT_STRUCTURE.md

@@ -25,7 +25,7 @@
 - `app/analysis/`
 - `app/web/`
 
-### 2. 运行主链路
+### 2. 运行服务链路
 
 - `app/services/altcoin_screener.py`
 - `app/services/altcoin_confirm.py`

static/app.html

@@ -890,7 +890,7 @@ function renderRecCard(r) {
       return '<span class="sig '+cls+'">'+displaySignalText(s)+'</span>';
     }).join('');
     var score = r.rec_score||0, st = scoreTier(score), ver = r.strategy_version||'';
-    var strategyLabel = r.strategy_name || ({main_composite_v1:'综合确认主链路',box_retest_1h_v1:'1H箱体突破回踩',box_retest_4h_v1:'4H箱体突破回踩'}[r.strategy_code||''] || r.strategy_code || '');
+    var strategyLabel = r.strategy_name || ({main_composite_v1:'综合确认策略',box_retest_1h_v1:'1H箱体突破回踩',box_retest_4h_v1:'4H箱体突破回踩',volume_ignition_1h_v1:'1H放量突破启动',compression_breakout_4h_v1:'4H压缩蓄力突破',intraday_momentum_15m_v1:'15m日内动量延续'}[r.strategy_code||''] || r.strategy_code || '');
     var hasQualityGate = ep.entry_quality_gate && Array.isArray(ep.entry_quality_gate.reasons) && ep.entry_quality_gate.reasons.length;
     var entryLabel = isWait ? '回踩参考' : (hasQualityGate ? '失效参考' : '参考价位');
     var entryRef = (isWait || hasQualityGate) ? (ep.entry_price || r.entry_price || 0) : (r.entry_price || ep.entry_price || 0);

static/onchain.html

@@ -53,7 +53,7 @@
       </div>
       <section class="panel">
         <div class="panel-head"><div class="panel-title">单币档案</div><div class="panel-note" id="detailNote">选择币种</div></div>
-        <div class="detail-body" id="detailBody"><div class="empty">从映射信号流或资产列表中选择一个币，查看链上事件与主链路关系。</div></div>
+        <div class="detail-body" id="detailBody"><div class="empty">从映射信号流或资产列表中选择一个币，查看链上事件与策略信号关系。</div></div>
       </section>
     </div>
     <div class="layout" style="grid-template-columns:1fr;margin-top:14px">
@@ -63,7 +63,7 @@
           <select class="select" id="chainSel" onchange="reloadTokens(0)"><option value="">全部链</option><option value="ethereum">Ethereum</option><option value="bsc">BSC</option></select>
           <select class="select" id="signalSel" onchange="reloadTokens(0)"><option value="">全部信号</option><option value="large_token_transfer">大额转账</option><option value="whale_accumulation">鲸鱼增持</option><option value="holder_growth">持有人增长</option><option value="exchange_outflow">交易所流出</option><option value="exchange_inflow_risk">交易所流入</option><option value="holder_concentration_risk">持仓集中风险</option></select>
         </div>
-        <div class="table-wrap"><table class="table"><thead><tr><th>币种</th><th>链</th><th>重要性</th><th>风险分</th><th>映射事件</th><th>最近事件</th><th>数据源</th><th>主链路</th></tr></thead><tbody id="tokenTable"><tr><td colspan="8" class="loading">加载中...</td></tr></tbody></table></div>
+        <div class="table-wrap"><table class="table"><thead><tr><th>币种</th><th>链</th><th>重要性</th><th>风险分</th><th>映射事件</th><th>最近事件</th><th>数据源</th><th>策略状态</th></tr></thead><tbody id="tokenTable"><tr><td colspan="8" class="loading">加载中...</td></tr></tbody></table></div>
         <div class="pager"><button id="prevBtn" onclick="page(-1)">上一页</button><span id="pageInfo">--</span><button id="nextBtn" onclick="page(1)">下一页</button></div>
       </section>
     </div>
@@ -94,7 +94,7 @@ async function reloadRawEvents(offset){state.rawOffset=offset||0;$('rawFeed').in
 async function reloadTokens(offset){state.offset=offset||0;$('tokenTable').innerHTML='<tr><td colspan="8" class="loading">加载中...</td></tr>';try{var qs='hours='+$('hoursSel').value+'&limit='+state.limit+'&offset='+state.offset+'&chain='+encodeURIComponent($('chainSel').value)+'&signal='+encodeURIComponent($('signalSel').value);var d=await (await fetch(API+'/api/onchain/tokens?'+qs)).json();state.total=d.total||0;var items=d.items||[];if(!items.length){$('tokenTable').innerHTML='<tr><td colspan="8" class="empty">暂无链上异动 token</td></tr>'}else{$('tokenTable').innerHTML=items.map(function(t){return '<tr onclick="loadDetail(\''+esc(t.symbol)+'\')"><td class="sym">'+esc(t.symbol)+'</td><td>'+esc(t.chain)+'</td><td class="num">'+Number(t.onchain_score||0).toFixed(0)+'</td><td class="num">'+Number(t.risk_score||0).toFixed(0)+'</td><td class="num">'+Number(t.event_count||t.mapped_event_count||0).toFixed(0)+'</td><td>'+esc(t.latest_event_at?fmtTime(t.latest_event_at):'--')+'</td><td>'+esc(t.source||'nodereal')+'</td><td>'+recLabel(t.recommendation)+'</td></tr>'}).join('');if(!state.selected&&items[0])loadDetail(items[0].symbol)}updatePager()}catch(e){$('tokenTable').innerHTML='<tr><td colspan="8" class="empty">加载失败</td></tr>'}}
 function updatePager(){var page=Math.floor(state.offset/state.limit)+1,totalPages=Math.max(1,Math.ceil((state.total||0)/state.limit));$('pageInfo').textContent='第 '+page+' / '+totalPages+' 页，共 '+state.total+' 个';$('prevBtn').disabled=state.offset<=0;$('nextBtn').disabled=state.offset+state.limit>=state.total}
 function page(step){var next=state.offset+step*state.limit;if(next<0||next>=state.total)return;reloadTokens(next)}
-async function loadDetail(symbol){state.selected=symbol;$('detailNote').textContent=symbol;$('detailBody').innerHTML='<div class="loading">加载详情...</div>';try{var d=await (await fetch(API+'/api/onchain/tokens/'+encodeURIComponent(symbol)+'?hours=168')).json();var latest=(d.metrics||[])[0]||{};var rec=d.recommendation;var rawCount=Number(d.raw_event_count||latest.event_count||latest.mapped_event_count||0);var metrics='<div class="metric-grid">'+[['重要性',Number(latest.onchain_score||0).toFixed(0)],['风险分',Number(latest.risk_score||0).toFixed(0)],['映射事件',rawCount.toFixed(0)],['数据源',latest.source||'nodereal']].map(function(x){return '<div class="metric"><span>'+x[0]+'</span><b>'+x[1]+'</b></div>'}).join('')+'</div>';var recHtml=rec?'<div class="hint">主链路状态：'+esc(rec.action_status||rec.execution_status||'观察')+' · 推荐 #'+esc(rec.id)+'</div>':'<div class="hint">尚未形成主链路推荐；若链上信号质量足够，会先进入技术检查。</div>';var standardEvents=(d.events||[]).slice(0,10).map(function(e){var cls=e.direction==='risk'?'risk':e.direction==='positive'?'pos':'blue';return '<div class="event"><div class="event-top"><div class="event-title">'+esc(e.signal_label||e.signal_code)+'</div><span class="badge '+cls+'">'+esc(e.severity||e.direction)+'</span></div><div class="event-meta">'+fmtTime(e.detected_at)+' · '+esc(e.chain)+' · '+fmtUsd(e.value_usd)+'<br>'+esc(e.wallet_label||e.counterparty_label||e.source||'链上事件')+'</div></div>'});var rawEvents=(d.raw_events||[]).slice(0,20).map(function(e){var amount=e.display_amount_label||fmtAmount(e.display_amount||e.total_amount||e.amount);var link=e.url?' · <a class="raw-link" href="'+esc(e.url)+'" target="_blank" rel="noopener">查看来源</a>':'';var route=(e.from_short&&e.to_short)?'<br>路径：'+esc(e.from_short)+' → '+esc(e.to_short):'';return '<div class="event"><div class="event-top"><div class="event-title">'+esc(e.human_summary||e.event_label||'NodeReal 原始事件')+'</div><span class="badge mapped">已映射</span></div><div class="event-meta">'+fmtTime(e.detected_at)+' · '+esc(e.chain)+' · '+esc(amount)+link+route+'<br>'+esc(e.pipeline_note||'已映射，可进入后续链上信号分析。')+'</div></div>'});var events=standardEvents.concat(rawEvents).join('')||'<div class="empty">暂无事件明细</div>';$('detailBody').innerHTML='<div class="detail-title">'+esc(d.symbol)+'</div><div class="detail-sub">'+(d.mappings||[]).length+' 个合约映射 · 近 7 天</div>'+metrics+recHtml+'<div class="event-feed">'+events+'</div>'}catch(e){$('detailBody').innerHTML='<div class="empty">详情加载失败</div>'}}
+async function loadDetail(symbol){state.selected=symbol;$('detailNote').textContent=symbol;$('detailBody').innerHTML='<div class="loading">加载详情...</div>';try{var d=await (await fetch(API+'/api/onchain/tokens/'+encodeURIComponent(symbol)+'?hours=168')).json();var latest=(d.metrics||[])[0]||{};var rec=d.recommendation;var rawCount=Number(d.raw_event_count||latest.event_count||latest.mapped_event_count||0);var metrics='<div class="metric-grid">'+[['重要性',Number(latest.onchain_score||0).toFixed(0)],['风险分',Number(latest.risk_score||0).toFixed(0)],['映射事件',rawCount.toFixed(0)],['数据源',latest.source||'nodereal']].map(function(x){return '<div class="metric"><span>'+x[0]+'</span><b>'+x[1]+'</b></div>'}).join('')+'</div>';var recHtml=rec?'<div class="hint">策略状态：'+esc(rec.action_status||rec.execution_status||'观察')+' · 推荐 #'+esc(rec.id)+'</div>':'<div class="hint">尚未形成策略推荐；若链上信号质量足够，会先进入技术检查。</div>';var standardEvents=(d.events||[]).slice(0,10).map(function(e){var cls=e.direction==='risk'?'risk':e.direction==='positive'?'pos':'blue';return '<div class="event"><div class="event-top"><div class="event-title">'+esc(e.signal_label||e.signal_code)+'</div><span class="badge '+cls+'">'+esc(e.severity||e.direction)+'</span></div><div class="event-meta">'+fmtTime(e.detected_at)+' · '+esc(e.chain)+' · '+fmtUsd(e.value_usd)+'<br>'+esc(e.wallet_label||e.counterparty_label||e.source||'链上事件')+'</div></div>'});var rawEvents=(d.raw_events||[]).slice(0,20).map(function(e){var amount=e.display_amount_label||fmtAmount(e.display_amount||e.total_amount||e.amount);var link=e.url?' · <a class="raw-link" href="'+esc(e.url)+'" target="_blank" rel="noopener">查看来源</a>':'';var route=(e.from_short&&e.to_short)?'<br>路径：'+esc(e.from_short)+' → '+esc(e.to_short):'';return '<div class="event"><div class="event-top"><div class="event-title">'+esc(e.human_summary||e.event_label||'NodeReal 原始事件')+'</div><span class="badge mapped">已映射</span></div><div class="event-meta">'+fmtTime(e.detected_at)+' · '+esc(e.chain)+' · '+esc(amount)+link+route+'<br>'+esc(e.pipeline_note||'已映射，可进入后续链上信号分析。')+'</div></div>'});var events=standardEvents.concat(rawEvents).join('')||'<div class="empty">暂无事件明细</div>';$('detailBody').innerHTML='<div class="detail-title">'+esc(d.symbol)+'</div><div class="detail-sub">'+(d.mappings||[]).length+' 个合约映射 · 近 7 天</div>'+metrics+recHtml+'<div class="event-feed">'+events+'</div>'}catch(e){$('detailBody').innerHTML='<div class="empty">详情加载失败</div>'}}
 function reloadAll(){state.offset=0;state.rawOffset=0;state.selected='';loadOverview();reloadRawEvents(0);reloadTokens(0)}
 reloadAll();
 setInterval(reloadAll,300000);

static/paper_trading.html

@@ -14,6 +14,9 @@
       <p>这里展示策略信号进入交易账本后的表现：只有系统把可买信号转入持仓或挂单后，才会进入收益统计。推荐历史和观察池不会直接产生收益率。</p>
     </div>
     <div class="actions">
+      <select class="select" id="strategyFilter" title="按策略筛选" onchange="onStrategyFilterChange()">
+        <option value="">全部策略</option>
+      </select>
       <button class="btn" id="sendReportBtn" onclick="sendReport()">发送策略交易报告</button>
       <button class="btn" onclick="loadAll()">刷新</button>
     </div>
@@ -132,7 +135,11 @@ function setTradeTab(tab){['open','orders','completed','events'].forEach(functio
 async function api(url){var r=await fetch(url);var d=await r.json().catch(function(){return{}});if(!r.ok)throw new Error(d.detail||d.error||'请求失败');return d}
 async function postApi(url){var r=await fetch(url,{method:'POST'});var d=await r.json().catch(function(){return{}});if(!r.ok)throw new Error(d.detail||d.error||'请求失败');return d}
 async function deleteApi(url){var r=await fetch(url,{method:'DELETE'});var d=await r.json().catch(function(){return{}});if(!r.ok)throw new Error((d.detail&&d.detail.reason)||d.detail||d.error||'请求失败');return d}
-async function loadAll(){await Promise.all([loadSummary(),loadPerformance(),loadOrders(),loadOpenTrades(openOffset),loadCompleted(),loadEvents(eventOffset)])}
+function selectedStrategy(){return $('strategyFilter')?($('strategyFilter').value||''):''}
+function strategyQuery(){var code=selectedStrategy();return code?'&strategy_code='+encodeURIComponent(code):''}
+async function loadAll(){await Promise.all([loadStrategies(),loadSummary(),loadPerformance(),loadOrders(),loadOpenTrades(openOffset),loadCompleted(),loadEvents(eventOffset)])}
+function onStrategyFilterChange(){openOffset=0;eventOffset=0;loadAll()}
+async function loadStrategies(){try{var d=await api('/api/paper-trading/strategies?days=120');var sel=$('strategyFilter');var current=sel.value||'';var rows=(d.strategies||[]).filter(function(x){return (x.signal_count||0)||(x.opportunity_count||0)||(x.trade_count||0)||(x.order_count||0)});sel.innerHTML='<option value="">全部策略</option>'+rows.map(function(x){return '<option value="'+esc(x.strategy_code)+'">'+esc(x.strategy_name||x.strategy_code)+' · '+esc(x.decision_label||'观察')+'</option>'}).join('');sel.value=Array.prototype.some.call(sel.options,function(o){return o.value===current})?current:''}catch(e){}}
 async function sendReport(){var btn=$('sendReportBtn'),note=$('reportNote');btn.disabled=true;btn.textContent='发送中...';note.style.display='block';note.textContent='正在汇总当前交易数据并发送飞书报告...';try{var d=await postApi('/api/paper-trading/report?days=30');note.textContent=d.ok?'报告已发送到飞书。':'报告生成完成，但飞书发送未成功：'+String(d.push_result||'未知原因');await loadSummary()}catch(e){note.textContent='发送失败：'+e.message}finally{btn.disabled=false;btn.textContent='发送策略交易报告'}}
 async function resetLedger(){var scope=$('resetScope').value||'all';var label=$('resetScope').selectedOptions[0]?$('resetScope').selectedOptions[0].textContent:scope;if(!confirm('确认重置“'+label+'”？这个操作会删除策略交易账本里的对应数据，不能从页面恢复。'))return;try{var d=await postApi('/api/paper-trading/reset?scope='+encodeURIComponent(scope));var del=d.deleted||{};alert('已重置：持仓/历史 '+(del.trades||0)+' 条，挂单 '+(del.orders||0)+' 条，日志 '+(del.events||0)+' 条。');await loadAll()}catch(e){alert('重置失败：'+e.message)}}
 async function deleteTrade(id,symbol,status){if(!confirm('确认删除 '+symbol+' 的'+(status==='open'?'持仓':'历史仓位')+'记录？相关操作日志也会一起删除。'))return;try{await deleteApi('/api/paper-trading/trades/'+encodeURIComponent(id));await loadAll()}catch(e){alert('删除失败：'+e.message)}}
@@ -153,7 +160,7 @@ function renderPerformance(d){var points=d.points||[];if(!points.length){$('perf
   '<span class="perf-pill">最大回撤 '+fmt(dd,2)+'%</span>',
   '<span class="perf-pill '+(pnl>=0?'pos':'neg')+'">总收益 '+(pnl>0?'+':'')+fmt(pnl,2)+'U</span>'
 ].join('');AlphaXCharts.renderEquity($('performanceChart'),d)}
-async function loadOrders(){$('orderRows').innerHTML='<tr><td colspan="11" class="loading">加载中...</td></tr>';try{var d=await api('/api/paper-trading/orders?limit=50&offset=0&status=pending');renderOrders(d.items||[])}catch(e){$('orderRows').innerHTML='<tr><td colspan="11" class="empty">'+esc(e.message)+'</td></tr>'}}
+async function loadOrders(){$('orderRows').innerHTML='<tr><td colspan="11" class="loading">加载中...</td></tr>';try{var d=await api('/api/paper-trading/orders?limit=50&offset=0&status=pending'+strategyQuery());renderOrders(d.items||[])}catch(e){$('orderRows').innerHTML='<tr><td colspan="11" class="empty">'+esc(e.message)+'</td></tr>'}}
 function renderOrders(items){if(!items.length){$('orderRows').innerHTML='<tr><td colspan="11" class="empty">暂无等待触价的策略挂单</td></tr>';return}$('orderRows').innerHTML=items.map(function(x){var latest=x.latest_price||x.current_price_at_create||0,dist=Number(x.distance_to_target_pct||0);return '<tr>'+
   '<td><div class="sym">'+esc(x.symbol)+'</div><div class="muted">#'+esc(x.id)+' · Rec '+esc(x.recommendation_id)+'</div></td>'+
   '<td><span class="badge open">'+esc(x.status==='pending'?'等待成交':x.status)+'</span></td>'+
@@ -168,11 +175,11 @@ function renderOrders(items){if(!items.length){$('orderRows').innerHTML='<tr><td
   '<td><button class="row-action" type="button" onclick="deleteOrder('+Number(x.id)+',\''+esc(String(x.symbol||'')).replace(/'/g,'&#39;')+'\')">删除</button></td>'+
   '</tr>'}).join('')}
 function protectionCell(x){var trail=Number(x.trailing_stop||0);var trailHtml=trail>0?'<span class="trail-line">移动止盈 $'+fmt(trail,6)+'</span>':'<span class="trail-line off">移动止盈未启动</span>';return '<div class="riskline"><span>TP $'+fmt(x.tp1,6)+'</span><span>SL $'+fmt(x.stop_loss,6)+'</span>'+trailHtml+'</div>'}
-function strategyName(x){return (x&&x.strategy_name)||({main_composite_v1:'综合确认主链路',box_retest_1h_v1:'1H箱体突破回踩',box_retest_4h_v1:'4H箱体突破回踩'}[(x&&x.strategy_code)||'']||((x&&x.strategy_code)||'--'))}
+function strategyName(x){return (x&&x.strategy_name)||({main_composite_v1:'综合确认策略',box_retest_1h_v1:'1H箱体突破回踩',box_retest_4h_v1:'4H箱体突破回踩',volume_ignition_1h_v1:'1H放量突破启动',compression_breakout_4h_v1:'4H压缩蓄力突破',intraday_momentum_15m_v1:'15m日内动量延续'}[(x&&x.strategy_code)||'']||((x&&x.strategy_code)||'--'))}
-async function loadOpenTrades(nextOffset){openOffset=Math.max(0,nextOffset||0);$('openRows').innerHTML='<tr><td colspan="14" class="loading">加载中...</td></tr>';try{var d=await api('/api/paper-trading/trades?limit='+LIMIT+'&offset='+openOffset+'&status=open');openTotal=d.total||0;renderTradeRows('openRows',d.items||[],'暂无持仓中的策略交易');renderOpenPager()}catch(e){$('openRows').innerHTML='<tr><td colspan="14" class="empty">'+esc(e.message)+'</td></tr>'}}
+async function loadOpenTrades(nextOffset){openOffset=Math.max(0,nextOffset||0);$('openRows').innerHTML='<tr><td colspan="14" class="loading">加载中...</td></tr>';try{var d=await api('/api/paper-trading/trades?limit='+LIMIT+'&offset='+openOffset+'&status=open'+strategyQuery());openTotal=d.total||0;renderTradeRows('openRows',d.items||[],'暂无持仓中的策略交易');renderOpenPager()}catch(e){$('openRows').innerHTML='<tr><td colspan="14" class="empty">'+esc(e.message)+'</td></tr>'}}
 async function loadCompleted(){await Promise.all([loadCompletedTrades(),loadCompletedOrders()])}
-async function loadCompletedTrades(){$('completedTradeRows').innerHTML='<tr><td colspan="14" class="loading">加载中...</td></tr>';try{var d=await api('/api/paper-trading/trades?limit=80&offset=0&status=closed');renderTradeRows('completedTradeRows',d.items||[],'暂无已平仓策略交易')}catch(e){$('completedTradeRows').innerHTML='<tr><td colspan="14" class="empty">'+esc(e.message)+'</td></tr>'}}
+async function loadCompletedTrades(){$('completedTradeRows').innerHTML='<tr><td colspan="14" class="loading">加载中...</td></tr>';try{var d=await api('/api/paper-trading/trades?limit=80&offset=0&status=closed'+strategyQuery());renderTradeRows('completedTradeRows',d.items||[],'暂无已平仓策略交易')}catch(e){$('completedTradeRows').innerHTML='<tr><td colspan="14" class="empty">'+esc(e.message)+'</td></tr>'}}
-async function loadCompletedOrders(){$('completedOrderRows').innerHTML='<tr><td colspan="11" class="loading">加载中...</td></tr>';try{var sets=await Promise.all(['filled','expired','canceled','rejected'].map(function(s){return api('/api/paper-trading/orders?limit=50&offset=0&status='+s)}));var items=[];sets.forEach(function(d){items=items.concat(d.items||[])});items.sort(function(a,b){return String(b.updated_at||b.created_at).localeCompare(String(a.updated_at||a.created_at))});renderCompletedOrders(items)}catch(e){$('completedOrderRows').innerHTML='<tr><td colspan="11" class="empty">'+esc(e.message)+'</td></tr>'}}
+async function loadCompletedOrders(){$('completedOrderRows').innerHTML='<tr><td colspan="11" class="loading">加载中...</td></tr>';try{var sets=await Promise.all(['filled','expired','canceled','rejected'].map(function(s){return api('/api/paper-trading/orders?limit=50&offset=0&status='+s+strategyQuery())}));var items=[];sets.forEach(function(d){items=items.concat(d.items||[])});items.sort(function(a,b){return String(b.updated_at||b.created_at).localeCompare(String(a.updated_at||a.created_at))});renderCompletedOrders(items)}catch(e){$('completedOrderRows').innerHTML='<tr><td colspan="11" class="empty">'+esc(e.message)+'</td></tr>'}}
 function renderTradeRows(targetId,items,emptyText){if(!items.length){$(targetId).innerHTML='<tr><td colspan="14" class="empty">'+esc(emptyText||'暂无策略交易')+'</td></tr>';return}$(targetId).innerHTML=items.map(function(x){var st=x.status==='open'?'持仓中':'已平仓';var latest=x.latest_price||x.current_price||0;var pnlUsdt=x.status==='closed'?x.realized_pnl_usdt:x.unrealized_pnl_usdt;return '<tr>'+
   '<td><div class="sym">'+esc(x.symbol)+'</div><div class="muted">#'+esc(x.id)+' · Rec '+esc(x.recommendation_id)+'</div></td>'+
   '<td><span class="badge '+esc(x.status)+'">'+st+'</span></td>'+
@@ -205,7 +212,7 @@ function renderCompletedOrders(items){if(!items.length){$('completedOrderRows').
 function orderStatus(x){return {filled:'已成交',expired:'已过期',canceled:'已取消',rejected:'已拒绝'}[x.status]||x.status||'--'}
 function cancelReasonLabel(r){return {global_risk_rejected:'全局风控拒绝：市场/账户风险过高，未转持仓',stop_loss_leverage_risk_exceeded:'止损杠杆风险超限：按当前止损和杠杆亏损过大',portfolio_drawdown_pause:'账户回撤保护：暂停新增仓位',weak_entries_pause:'近期弱入场过多：暂停新增仓位',recommendation_invalid:'原推荐已失效：机会过期/归档后撤单',too_far_from_entry:'价格远离计划价：继续等待意义不大',expired:'挂单超时：超过有效期未成交',upgraded_to_buy_now:'信号升级为入场窗口：旧挂单已撤销',canceled:'已取消',filled:'已成交',rejected:'已拒绝'}[r]||r||'--'}
 function renderOpenPager(){var page=Math.floor(openOffset/LIMIT)+1,totalPages=Math.max(1,Math.ceil(openTotal/LIMIT));$('openPageInfo').textContent='第 '+page+' / '+totalPages+' 页 · 共 '+openTotal+' 条';$('openPager').innerHTML='<button '+(openOffset===0?'disabled':'')+' onclick="loadOpenTrades('+(openOffset-LIMIT)+')">上一页</button><span>第 '+page+' / '+totalPages+' 页</span><button '+((openOffset+LIMIT>=openTotal)?'disabled':'')+' onclick="loadOpenTrades('+(openOffset+LIMIT)+')">下一页</button>'}
-async function loadEvents(nextOffset){eventOffset=Math.max(0,nextOffset||0);$('eventRows').innerHTML='<div class="loading">加载中...</div>';try{var sym=$('eventSymbol').value||'';var typ=$('eventType').value||'';var d=await api('/api/paper-trading/events?limit='+EVENT_LIMIT+'&offset='+eventOffset+'&symbol='+encodeURIComponent(sym)+'&event_type='+encodeURIComponent(typ));eventTotal=d.total||0;renderEvents(d.items||[]);renderEventPager()}catch(e){$('eventRows').innerHTML='<div class="empty">'+esc(e.message)+'</div>'}}
+async function loadEvents(nextOffset){eventOffset=Math.max(0,nextOffset||0);$('eventRows').innerHTML='<div class="loading">加载中...</div>';try{var sym=$('eventSymbol').value||'';var typ=$('eventType').value||'';var d=await api('/api/paper-trading/events?limit='+EVENT_LIMIT+'&offset='+eventOffset+'&symbol='+encodeURIComponent(sym)+'&event_type='+encodeURIComponent(typ)+strategyQuery());eventTotal=d.total||0;renderEvents(d.items||[]);renderEventPager()}catch(e){$('eventRows').innerHTML='<div class="empty">'+esc(e.message)+'</div>'}}
 function eventLabel(t){return {open:'开仓',close:'平仓',trailing_activate:'移动止盈激活',trailing_move:'移动止盈上移'}[t]||t||'动作'}
 function eventCls(t){if(t==='open')return'event-open';if(t==='close')return'event-close';if(String(t).indexOf('trailing')===0)return'event-trailing';return''}
 function contextPill(text,cls){return '<span class="ctx-pill '+esc(cls||'')+'">'+esc(text)+'</span>'}

tests/test_llm_insights.py

@@ -79,7 +79,7 @@ def _fetch_llm_row(db_path):
     return dict(row) if row else None
 
 
-def test_disabled_llm_skips_and_does_not_change_mainline(monkeypatch, temp_db):
+def test_disabled_llm_skips_and_does_not_change_strategy_state(monkeypatch, temp_db):
     _insert_recommendation(temp_db)
     monkeypatch.setattr(llm_insights, "get_llm_module_enabled", lambda module: False)
     result = llm_insights.run(scope="recommendations", limit=10)

tests/test_multi_strategy_infra.py

@@ -1,9 +1,23 @@
 from app.core.factor_roles import RISK, TRIGGER, factor_role, factor_roles_for_codes
 from app.core.strategy_contract import StrategySignal, default_main_composite_signal
-from app.core.strategy_registry import BOX_RETEST_1H_STRATEGY, BOX_RETEST_4H_STRATEGY, MAIN_COMPOSITE_STRATEGY, strategy_label
+from app.core.strategy_registry import (
+    BOX_RETEST_1H_STRATEGY,
+    BOX_RETEST_4H_STRATEGY,
+    COMPRESSION_BREAKOUT_4H_STRATEGY,
+    INTRADAY_MOMENTUM_15M_STRATEGY,
+    MAIN_COMPOSITE_STRATEGY,
+    VOLUME_IGNITION_1H_STRATEGY,
+    strategy_label,
+)
 from app.db.recommendation_commands import create_recommendation
-from app.db.strategy_signal_queries import insert_strategy_signal
 from app.db.paper_trading import _open_trade, _order_payload_from_rec
+from app.db.strategy_signal_queries import insert_strategy_signal
+from app.db.strategy_insights import evaluate_strategy_decision
+from app.strategies.altcoin_breakout import (
+    build_compression_breakout_4h_signal,
+    build_intraday_momentum_15m_signal,
+    build_volume_ignition_1h_signal,
+)
 from app.strategies.box_retest_4h import build_box_retest_1h_signal
 
 
@@ -27,9 +41,109 @@ def test_default_main_composite_strategy_signal_is_stable():
     ).to_json_dict()
 
     assert signal["strategy_code"] == MAIN_COMPOSITE_STRATEGY
-    assert signal["strategy_name"] == "综合确认主链路"
+    assert signal["strategy_name"] == "综合确认策略"
     assert signal["factor_roles"]["vp_fly_1h_current"] == "trigger"
     assert strategy_label(BOX_RETEST_1H_STRATEGY) == "1H箱体突破回踩"
+    assert strategy_label(VOLUME_IGNITION_1H_STRATEGY) == "1H放量突破启动"
+    assert strategy_label(COMPRESSION_BREAKOUT_4H_STRATEGY) == "4H压缩蓄力突破"
+    assert strategy_label(INTRADAY_MOMENTUM_15M_STRATEGY) == "15m日内动量延续"
+
+
+def test_volume_ignition_strategy_builds_independent_signal():
+    signal = build_volume_ignition_1h_signal(
+        symbol="VOL/USDT",
+        result={
+            "score": 8,
+            "signals": ["1H量价齐飞 · 连续放量", "15min即刻入场"],
+            "trigger_context": {"current_triggers": ["15m突破"], "trigger_status": "current"},
+            "entry_plan": {"entry_action": "可即刻买入"},
+        },
+        entry_plan={"entry_action": "可即刻买入", "entry_price": 1.0},
+    )
+    payload = signal.to_json_dict()
+
+    assert payload["strategy_code"] == VOLUME_IGNITION_1H_STRATEGY
+    assert payload["status"] == "candidate"
+    assert payload["trigger"]["factor_code"] == "vp_fly_1h_current"
+    assert payload["factor_roles"]["vp_fly_1h_current"] == "trigger"
+
+
+def test_compression_breakout_strategy_requires_structure_and_breakout_context():
+    signal = build_compression_breakout_4h_signal(
+        symbol="QUIET/USDT",
+        result={"score": 6, "signals": ["4H静K压缩，突破箱体上沿"], "entry_plan": {"entry_action": "等回踩"}},
+        entry_plan={"entry_action": "等回踩", "entry_price": 1.0},
+    )
+    payload = signal.to_json_dict()
+
+    assert payload["strategy_code"] == COMPRESSION_BREAKOUT_4H_STRATEGY
+    assert payload["status"] == "candidate"
+    assert payload["factor_roles"]["compression_surge_4h"] == "trigger"
+    assert build_compression_breakout_4h_signal(
+        symbol="NOBOX/USDT",
+        result={"score": 6, "signals": ["1H量价齐飞"], "entry_plan": {"entry_action": "可即刻买入"}},
+        entry_plan={"entry_action": "可即刻买入"},
+    ) is None
+
+
+def test_intraday_momentum_strategy_requires_current_trigger():
+    stale = build_intraday_momentum_15m_signal(
+        symbol="FAST/USDT",
+        result={"score": 7, "signals": ["15m短周期启动"], "entry_plan": {"entry_action": "可即刻买入"}},
+        entry_plan={"entry_action": "可即刻买入"},
+    ).to_json_dict()
+    fresh = build_intraday_momentum_15m_signal(
+        symbol="FAST/USDT",
+        result={
+            "score": 7,
+            "signals": ["15min强突破"],
+            "trigger_context": {"current_triggers": ["15m突破"], "trigger_status": "current"},
+            "entry_plan": {"entry_action": "可即刻买入"},
+        },
+        entry_plan={"entry_action": "可即刻买入"},
+    ).to_json_dict()
+
+    assert stale["status"] == "observe"
+    assert "缺少当前低周期触发" in stale["risk_plan"]["risk_reasons"]
+    assert fresh["status"] == "candidate"
+
+
+def test_strategy_evaluation_recommends_promote_or_pause():
+    strong = evaluate_strategy_decision({
+        "signal_count": 24,
+        "opportunity_count": 16,
+        "trade_count": 8,
+        "closed_trade_count": 8,
+        "win_rate_pct": 62.5,
+        "avg_realized_pnl_pct": 3.2,
+        "realized_pnl_usdt": 180,
+        "worst_pnl_pct": -3.5,
+        "order_fill_rate_pct": 45,
+        "trade_conversion_pct": 50,
+    })
+    weak = evaluate_strategy_decision({
+        "signal_count": 24,
+        "opportunity_count": 16,
+        "trade_count": 8,
+        "closed_trade_count": 8,
+        "win_rate_pct": 25,
+        "avg_realized_pnl_pct": -2.5,
+        "realized_pnl_usdt": -120,
+        "worst_pnl_pct": -9,
+        "order_fill_rate_pct": 20,
+        "trade_conversion_pct": 50,
+    })
+    unfilled = evaluate_strategy_decision({
+        "signal_count": 18,
+        "opportunity_count": 12,
+        "trade_count": 0,
+        "closed_trade_count": 0,
+    })
+
+    assert strong["decision"] == "promote"
+    assert strong["evaluation_score"] > weak["evaluation_score"]
+    assert weak["decision"] == "pause"
+    assert unfilled["decision"] == "review_entry_gate"
 
 
 def test_strategy_signal_insert_and_recommendation_lineage(pg_conn):