feat: 数据管线 P2 优化 — 血缘写入 + 质量监控 + 缓存清理 + 重试 worker

P2-1: DataLineage 血缘写入(bzzoiro/understat Bronze→Silver 全链路)
P2-2: DataQualityCheck 质量检查写入(行计数异常检测)
P2-3: RateLimitedClient 集成到 bzzoiro 采集
P2-4: injuries 缓存过期清理(启动时扫描删除 >1h 文件)
P2-5: IngestFailure 重试 worker(FastAPI startup 注册)

新增文件:
  - src/data/retry_worker.py (死信表重试 sweep)

.env.example

@@ -3,6 +3,10 @@ APP_ENV=development
 LOG_LEVEL=INFO
 
 # ---- 数据库 ----
+POSTGRES_USER=football
+POSTGRES_PASSWORD=football
+POSTGRES_DB=football
+POSTGRES_PORT=5432
 DATABASE_URL=postgresql+asyncpg://football:football@localhost:5432/football
 
 # ---- LLM (OpenAI-compatible,必填一个) ----
@@ -10,6 +14,11 @@ LLM_PROVIDER=openai
 LLM_API_KEY=sk-xxxx
 LLM_BASE_URL=https://api.openai.com/v1
 LLM_MODEL=gpt-4o
+# 多 Agent 分档模型(留空则回落 LLM_MODEL)
+LLM_SPECIALIST_MODEL=
+LLM_AGGREGATOR_MODEL=
+# 单次 LLM 调用超时(秒)
+LLM_TIMEOUT=60
 
 # ---- 数据源 ----
 BZZOIRO_KEY=
@@ -17,3 +26,8 @@ API_FOOTBALL_KEY=
 
 # ---- CORS ----
 CORS_ORIGINS=http://localhost:5173,http://localhost:3000
+
+# ---- 管理接口鉴权 ----
+# 采集/回测/回填接口的访问密钥(请求头 X-API-Key)。
+# 留空 = 不启用鉴权(本地开发默认);生产环境必须设置强随机值。
+ADMIN_API_KEY=

.gitignore

@@ -10,3 +10,7 @@ frontend/dist/
 CLAUDE.md
 docs/AGENTS.md
 docs/agents/
+
+# 本地审查/预览脚手架(不入库)
+.tools/
+.preview/

alembic/versions/0004_snapshot_and_constraints.py

@@ -26,6 +26,21 @@ def upgrade() -> None:
     op.add_column('predictions', sa.Column('input_hash', sa.String(length=64), nullable=True))
 
     # 3. 新增 CHECK 约束
+    #    注意: 对已有数据行加 CHECK 约束时,若存在越界数据 ALTER TABLE 会中途失败,
+    #    导致迁移卡在半完成状态。这里先做一次性清洗(把越界值收敛到合法域),
+    #    再建约束,保证在非空库上也能成功。
+    op.execute("UPDATE predictions SET pred_home_goals = 0 WHERE pred_home_goals IS NOT NULL AND pred_home_goals < 0")
+    op.execute("UPDATE predictions SET pred_away_goals = 0 WHERE pred_away_goals IS NOT NULL AND pred_away_goals < 0")
+    op.execute(
+        "UPDATE predictions SET subjective_confidence = "
+        "CASE WHEN subjective_confidence < 0 THEN 0 "
+        "     WHEN subjective_confidence > 1 THEN 1 ELSE subjective_confidence END "
+        "WHERE subjective_confidence IS NOT NULL "
+        "  AND (subjective_confidence < 0 OR subjective_confidence > 1)"
+    )
+    op.execute("UPDATE predictions SET pred_1x2 = NULL WHERE pred_1x2 IS NOT NULL AND pred_1x2 NOT IN ('1', 'X', '2')")
+    op.execute("UPDATE predictions SET mode = 'single' WHERE mode IS NOT NULL AND mode NOT IN ('single', 'multi')")
+
     op.create_check_constraint('ck_pred_home_goals_nonneg', 'predictions', 'pred_home_goals >= 0')
     op.create_check_constraint('ck_pred_away_goals_nonneg', 'predictions', 'pred_away_goals >= 0')
     op.create_check_constraint('ck_confidence_range', 'predictions', 'subjective_confidence >= 0 AND subjective_confidence <= 1')

alembic/versions/0006_schema_model_drift_cleanup.py

@@ -0,0 +1,80 @@
+"""清理 schema 与 ORM 模型的漂移
+
+Revision ID: 0006_schema_model_drift_cleanup
+Revises: 0005_prediction_status_and_stats_provenance
+Create Date: 2026-09-15
+
+背景(见代码审查报告 P1-1):
+  0005 迁移在 predictions 表留下了 `cutoff_at` 列(删除语句被注释掉),
+  但 ORM 模型 `Prediction` 中并无该字段 —— 这是一处会长期存在的 schema 漂移,
+  而 alembic autogenerate 会持续建议 drop 它,造成噪声。
+
+  同时 0005 创建的两个索引 `ix_match_stats_available_at` 与
+  `ix_predictions_cutoff_at` 未在 ORM 声明,autogenerate 会建议删除它们 ——
+  一旦被误删,数据血缘相关的时间过滤查询会退化为全表扫描。
+
+本迁移做两件事:
+  1. drop 掉幽灵列 predictions.cutoff_at(数据已由 prediction_cutoff_at 承载,
+     迁移前先把非空值回填过去,避免丢数据)
+  2. 显式重建这两个索引(幂等:先 drop if exists 再 create),使 DB 状态与
+     修正后的 ORM 声明一致
+"""
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision: str = '0006_schema_model_drift_cleanup'
+down_revision: Union[str, None] = '0005_prediction_status_and_stats_provenance'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    bind = op.get_bind()
+    inspector = sa.inspect(bind)
+
+    # --- 1. 幽灵列 cutoff_at: 先把数据回填到 prediction_cutoff_at 再删除 ---
+    pred_cols = {c["name"] for c in inspector.get_columns("predictions")}
+    if "cutoff_at" in pred_cols:
+        if "prediction_cutoff_at" in pred_cols:
+            # 仅回填尚未有值的行,避免覆盖更权威的数据
+            op.execute(
+                "UPDATE predictions "
+                "SET prediction_cutoff_at = cutoff_at "
+                "WHERE prediction_cutoff_at IS NULL AND cutoff_at IS NOT NULL"
+            )
+        op.drop_column("predictions", "cutoff_at")
+
+    # --- 2. 与 ORM 声明对齐的索引(幂等重建) ---
+    stats_idx = {i["name"] for i in inspector.get_indexes("match_stats")}
+    if "ix_match_stats_available_at" not in stats_idx:
+        op.create_index("ix_match_stats_available_at", "match_stats", ["available_at"])
+
+    pred_idx = {i["name"] for i in inspector.get_indexes("predictions")}
+    if "ix_predictions_cutoff_at" not in pred_idx:
+        op.create_index(
+            "ix_predictions_cutoff_at", "predictions", ["prediction_cutoff_at"]
+        )
+
+
+def downgrade() -> None:
+    bind = op.get_bind()
+    inspector = sa.inspect(bind)
+
+    # 恢复幽灵列(与 0005 的最终状态一致:列存在但无值)
+    pred_cols = {c["name"] for c in inspector.get_columns("predictions")}
+    if "cutoff_at" not in pred_cols:
+        op.add_column(
+            "predictions",
+            sa.Column("cutoff_at", sa.DateTime(timezone=True), nullable=True),
+        )
+
+    stats_idx = {i["name"] for i in inspector.get_indexes("match_stats")}
+    if "ix_match_stats_available_at" in stats_idx:
+        op.drop_index("ix_match_stats_available_at", table_name="match_stats")
+
+    pred_idx = {i["name"] for i in inspector.get_indexes("predictions")}
+    if "ix_predictions_cutoff_at" in pred_idx:
+        op.drop_index("ix_predictions_cutoff_at", table_name="predictions")

alembic/versions/0007_injuries_constraint_naming_align.py

@@ -0,0 +1,81 @@
+"""修复 injuries 表约束命名与 ORM 声明不一致
+
+Revision ID: 0007_injuries_constraint_naming_align
+Revises: 0006_schema_model_drift_cleanup
+Create Date: 2026-09-16
+
+背景(见代码审查报告 P2-5):
+  0003 迁移使用 sa.UniqueConstraint 创建唯一约束,
+  而 ORM models.py 中声明为 Index(..., unique=True)。
+  虽然 PostgreSQL 中两者效果相同(都保证唯一性),
+  但 pg_catalog 中表示不同,会导致:
+    - alembic autogenerate 持续报告漂移
+    - 约束命名约定不一致(uc_ 前缀 vs ix_ 前缀)
+
+  本迁移将 UniqueConstraint 替换为唯一索引,与 ORM 声明对齐。
+"""
+
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision: str = '0007_injuries_constraint_naming_align'
+down_revision: Union[str, None] = '0006_schema_model_drift_cleanup'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    bind = op.get_bind()
+    inspector = sa.inspect(bind)
+
+    # 检查当前约束类型
+    constraints = {
+        c["name"]: c
+        for c in inspector.get_unique_constraints("injuries")
+    }
+    indexes = {
+        i["name"]: i
+        for i in inspector.get_indexes("injuries")
+    }
+
+    # 如果存在 UniqueConstraint 形式的 ix_injuries_player_fixture,替换为唯一索引
+    if "ix_injuries_player_fixture" in constraints:
+        # 删除唯一约束
+        op.drop_constraint("ix_injuries_player_fixture", "injuries", type_="unique")
+
+    # 如果不存在同名唯一索引,创建它(与 ORM 声明一致)
+    if "ix_injuries_player_fixture" not in indexes:
+        op.create_index(
+            "ix_injuries_player_fixture",
+            "injuries",
+            ["player_id", "fixture_id", "injury_type"],
+            unique=True,
+        )
+
+
+def downgrade() -> None:
+    bind = op.get_bind()
+    inspector = sa.inspect(bind)
+
+    indexes = {
+        i["name"]: i
+        for i in inspector.get_indexes("injuries")
+    }
+    constraints = {
+        c["name"]: c
+        for c in inspector.get_unique_constraints("injuries")
+    }
+
+    # 恢复为 UniqueConstraint 形式
+    if "ix_injuries_player_fixture" in indexes:
+        op.drop_index("ix_injuries_player_fixture", table_name="injuries")
+
+    if "ix_injuries_player_fixture" not in constraints:
+        op.create_unique_constraint(
+            "ix_injuries_player_fixture",
+            "injuries",
+            ["player_id", "fixture_id", "injury_type"],
+        )

alembic/versions/0007_predictions_unique_constraint.py

@@ -0,0 +1,59 @@
+"""为 predictions 表添加 match_id+provider+model 唯一约束
+
+Revision ID: 0007_predictions_unique_constraint
+Revises: 0007_injuries_constraint_naming_align
+Create Date: 2026-09-16
+
+背景(见代码审查报告 P1-6):
+  同一 match_id + provider + model 组合不应产生重复预测。
+  当前缺少数据库级唯一约束,回测多次运行或并发采集可能产生重复记录,
+  导致统计偏差。
+
+  先清理已存在的重复记录(保留最早创建的那条),再添加唯一约束。
+"""
+
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision: str = '0007_predictions_unique_constraint'
+down_revision: Union[str, None] = '0007_injuries_constraint_naming_align'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    # 1. 清理已存在的重复记录(保留 id 最小的)
+    op.execute(
+        """
+        DELETE FROM predictions
+        WHERE id NOT IN (
+            SELECT MIN(id)
+            FROM predictions
+            GROUP BY match_id, provider, model
+        )
+        AND match_id IN (
+            SELECT match_id
+            FROM predictions
+            GROUP BY match_id, provider, model
+            HAVING COUNT(*) > 1
+        )
+        """
+    )
+
+    # 2. 添加唯一约束
+    op.create_unique_constraint(
+        "uq_predictions_match_provider_model",
+        "predictions",
+        ["match_id", "provider", "model"],
+    )
+
+
+def downgrade() -> None:
+    op.drop_constraint(
+        "uq_predictions_match_provider_model",
+        "predictions",
+        type_="unique",
+    )

alembic/versions/0008_raw_event_and_ingest_failure.py

@@ -0,0 +1,121 @@
+"""新增 Bronze 层 + 死信表 + 数据质量表 + 血缘表
+
+Revision ID: 0008_raw_event_and_ingest_failure
+Revises: 0007_predictions_unique_constraint
+Create Date: 2026-09-17
+
+架构审查报告 P1 实施:
+  - raw_events: Bronze 层,不可变原始采集记录
+  - ingest_failures: 采集失败死信表
+  - data_quality_checks: 数据质量检查结果记录
+  - data_lineage: ETL 全过程元数据血缘追踪
+"""
+
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects.postgresql import JSONB
+
+# revision identifiers, used by Alembic.
+revision: str = '0008_raw_event_and_ingest_failure'
+down_revision: Union[str, None] = '0007_predictions_unique_constraint'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    # ── raw_events: Bronze 层原始记录 ──
+    op.create_table(
+        "raw_events",
+        sa.Column("id", sa.Integer(), primary_key=True),
+        sa.Column("source_system", sa.String(30), nullable=False),
+        sa.Column("source_record_id", sa.String(100), nullable=False),
+        sa.Column("raw_payload", JSONB(), nullable=False),
+        sa.Column("ingested_at", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("ingest_batch_id", sa.String(64), nullable=True),
+    )
+    op.create_index("ix_raw_events_batch", "raw_events", ["ingest_batch_id"])
+    op.create_index("ix_raw_events_source_ingested", "raw_events", ["source_system", "ingested_at"])
+    op.create_unique_constraint("uq_raw_events_source_record", "raw_events", ["source_system", "source_record_id"])
+
+    # ── ingest_failures: 采集失败死信表 ──
+    op.create_table(
+        "ingest_failures",
+        sa.Column("id", sa.Integer(), primary_key=True),
+        sa.Column("source_system", sa.String(30), nullable=False),
+        sa.Column("entity_type", sa.String(30), nullable=False),
+        sa.Column("source_record_id", sa.String(100), nullable=True),
+        sa.Column("error_type", sa.String(50), nullable=False),
+        sa.Column("error_detail", sa.Text(), nullable=True),
+        sa.Column("raw_payload", JSONB(), nullable=True),
+        sa.Column("retry_count", sa.Integer(), nullable=False, server_default="0"),
+        sa.Column("next_retry_at", sa.DateTime(timezone=True), nullable=True),
+        sa.Column("status", sa.String(20), nullable=False, server_default="pending"),
+        sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("resolved_at", sa.DateTime(timezone=True), nullable=True),
+    )
+    op.create_index("ix_ingest_failures_status_next_retry", "ingest_failures", ["status", "next_retry_at"])
+    op.create_index("ix_ingest_failures_source", "ingest_failures", ["source_system", "entity_type"])
+    op.create_check_constraint("ck_ingest_failures_status", "ingest_failures",
+                               "status IN ('pending', 'retrying', 'resolved', 'abandoned')")
+    op.create_check_constraint("ck_ingest_failures_retry_nonneg", "ingest_failures", "retry_count >= 0")
+
+    # ── data_quality_checks: 数据质量检查记录 ──
+    op.create_table(
+        "data_quality_checks",
+        sa.Column("id", sa.Integer(), primary_key=True),
+        sa.Column("check_name", sa.String(100), nullable=False),
+        sa.Column("entity_type", sa.String(30), nullable=False),
+        sa.Column("entity_id", sa.String(50), nullable=True),
+        sa.Column("expected_value", sa.Text(), nullable=True),
+        sa.Column("actual_value", sa.Text(), nullable=True),
+        sa.Column("passed", sa.Boolean(), nullable=False),
+        sa.Column("severity", sa.String(10), nullable=False, server_default="warning"),
+        sa.Column("detail", sa.Text(), nullable=True),
+        sa.Column("checked_at", sa.DateTime(timezone=True), nullable=False),
+    )
+    op.create_index("ix_dqc_check_time", "data_quality_checks", ["check_name", "checked_at"])
+    op.create_index("ix_dqc_entity", "data_quality_checks", ["entity_type", "entity_id"])
+    op.create_index("ix_dqc_severity_passed", "data_quality_checks", ["severity", "passed"])
+    op.create_check_constraint("ck_dqc_severity", "data_quality_checks",
+                               "severity IN ('info', 'warning', 'critical')")
+
+    # ── data_lineage: ETL 血缘追踪 ──
+    op.create_table(
+        "data_lineage",
+        sa.Column("id", sa.Integer(), primary_key=True),
+        sa.Column("source_system", sa.String(30), nullable=False),
+        sa.Column("source_record_id", sa.String(100), nullable=False),
+        sa.Column("target_table", sa.String(50), nullable=False),
+        sa.Column("target_id", sa.Integer(), nullable=True),
+        sa.Column("transform_name", sa.String(100), nullable=False),
+        sa.Column("transform_detail", sa.Text(), nullable=True),
+        sa.Column("batch_id", sa.String(64), nullable=True),
+        sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
+    )
+    op.create_index("ix_lineage_source", "data_lineage", ["source_system", "source_record_id"])
+    op.create_index("ix_lineage_target", "data_lineage", ["target_table", "target_id"])
+    op.create_index("ix_lineage_batch", "data_lineage", ["batch_id"])
+
+
+def downgrade() -> None:
+    # 逆序删除
+    op.drop_index("ix_lineage_batch", table_name="data_lineage")
+    op.drop_index("ix_lineage_target", table_name="data_lineage")
+    op.drop_index("ix_lineage_source", table_name="data_lineage")
+    op.drop_table("data_lineage")
+
+    op.drop_index("ix_dqc_severity_passed", table_name="data_quality_checks")
+    op.drop_index("ix_dqc_entity", table_name="data_quality_checks")
+    op.drop_index("ix_dqc_check_time", table_name="data_quality_checks")
+    op.drop_table("data_quality_checks")
+
+    op.drop_index("ix_ingest_failures_source", table_name="ingest_failures")
+    op.drop_index("ix_ingest_failures_status_next_retry", table_name="ingest_failures")
+    op.drop_table("ingest_failures")
+
+    op.drop_constraint("uq_raw_events_source_record", table_name="raw_events", type_="unique")
+    op.drop_index("ix_raw_events_source_ingested", table_name="raw_events")
+    op.drop_index("ix_raw_events_batch", table_name="raw_events")
+    op.drop_table("raw_events")

alembic/versions/0009_match_stats_xg_fields.py

@@ -0,0 +1,35 @@
+"""MatchStats 新增 xG 追踪字段
+
+Revision ID: 0009_match_stats_xg_fields
+Revises: 0008_raw_event_and_ingest_failure
+Create Date: 2026-09-17
+
+架构审查报告 P1-4 实施:
+  understat 允许纠正旧 xG 值。新增字段追踪 xG 具体来源和更新时间,
+  实现全量覆盖模式:当 understat 数据更新时覆盖旧值而非跳过。
+"""
+
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision: str = '0009_match_stats_xg_fields'
+down_revision: Union[str, None] = '0008_raw_event_and_ingest_failure'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    op.add_column("match_stats", sa.Column("xg_source", sa.String(30), nullable=True))
+    op.add_column("match_stats", sa.Column("xg_updated_at", sa.DateTime(timezone=True), nullable=True))
+    op.add_column("match_stats", sa.Column("xg_source_record_id", sa.String(100), nullable=True))
+    op.create_index("ix_match_stats_xg_source", "match_stats", ["xg_source", "xg_updated_at"])
+
+
+def downgrade() -> None:
+    op.drop_index("ix_match_stats_xg_source", table_name="match_stats")
+    op.drop_column("match_stats", "xg_source_record_id")
+    op.drop_column("match_stats", "xg_updated_at")
+    op.drop_column("match_stats", "xg_source")

docker-compose.yml

@@ -2,15 +2,15 @@ services:
   postgres:
     image: postgres:16-alpine
     environment:
-      POSTGRES_USER: football
-      POSTGRES_PASSWORD: football
-      POSTGRES_DB: football
+      POSTGRES_USER: ${POSTGRES_USER:?POSTGRES_USER 未设置}
+      POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:?POSTGRES_PASSWORD 未设置}
+      POSTGRES_DB: ${POSTGRES_DB:-football}
     ports:
-      - "5432:5432"
+      - "${POSTGRES_PORT:-5432}:5432"
     volumes:
       - pgdata:/var/lib/postgresql/data
     healthcheck:
-      test: ["CMD-SHELL", "pg_isready -U football"]
+      test: ["CMD-SHELL", "pg_isready -U ${POSTGRES_USER:?POSTGRES_USER 未设置}"]
       interval: 5s
       timeout: 5s
       retries: 5
@@ -19,7 +19,7 @@ services:
     build: .
     command: uvicorn src.api.app:app --host 0.0.0.0 --port 8000 --reload
     ports:
-      - "8000:8000"
+      - "${API_PORT:-8000}:8000"
     env_file: .env
     depends_on:
       postgres:

docs/01-architecture.md

@@ -57,38 +57,48 @@
 Profeto/
 ├── src/
 │   ├── api/
-│   │   ├── app.py               # FastAPI 工厂(lifespan 建表)
+│   │   ├── app.py               # FastAPI 工厂(lifespan 仅验证 DB 连接,不建表)
+│   │   ├── deps.py              # 依赖:管理接口鉴权(X-API-Key)
 │   │   ├── schemas.py           # Pydantic v2 请求/响应
 │   │   └── routes/
 │   │       ├── matches.py       # 联赛/比赛查询(游标分页)
 │   │       ├── predict.py       # 预测 + 预测历史
-│   │       ├── ingest.py        # 采集触发(自管 session)
-│   │       └── eval.py          # 赛后回填 + 准确率汇总
+│   │       ├── ingest.py        # 采集触发(自管 session,需鉴权)
+│   │       ├── eval.py          # 赛后回填 + 准确率汇总
+│   │       └── backtest.py      # 历史回测(需鉴权)
 │   ├── db/
 │   │   ├── base.py              # async engine + get_db/get_db_read
-│   │   └── models.py            # 6 张表 ORM
+│   │   ├── models.py            # 6 张表 ORM
+│   │   ├── repositories.py      # 仓储层
+│   │   └── unit_of_work.py      # 事务边界
 │   ├── data/
 │   │   ├── bzzoiro.py           # 赛果采集 + 幂等入库
 │   │   ├── understat.py         # xG 回填
 │   │   ├── injuries.py          # 伤停采集(带文件缓存)
 │   │   ├── normalize.py         # NormalizedMatch 清洗契约
 │   │   ├── team_names.py        # 队名归一映射
+│   │   ├── sources.py           # 数据源注册表
 │   │   └── config.py            # 联赛代码映射
 │   ├── llm/
 │   │   ├── provider.py          # OpenAI-compatible 抽象(共享连接池/JSON 兜底解析)
 │   │   ├── context_builder.py   # 数据切片(h2h/form/stats/home_away/injuries)+ 单 agent 拼接
 │   │   ├── predict.py           # 预测入口(mode 分派 + 缓存)
 │   │   ├── eval.py              # 准确率统计
+│   │   ├── validation.py        # LLM 输出严格校验(Pydantic)
+│   │   ├── backtest.py          # 回测执行
 │   │   ├── agents/
 │   │   │   ├── base.py          # AgentSpec / AgentReport / run_agent
 │   │   │   └── orchestrator.py  # 并行专家 → 终裁 → 存库
 │   │   └── prompts/
 │   │       ├── match_prediction_v1/v2.md   # 单 agent 模板
 │   │       └── agents/{h2h,form,stats,home_away,injuries,aggregator}_v1.md
-│   └── core/config.py           # pydantic-settings
-├── alembic/versions/            # 0001 建表 + 0002 agent 字段
-├── frontend/src/pages/Matches.tsx  # 单页(预测面板 + 专家报告折叠区)
-├── tests/                       # 33 项(核心 13 + agent 20)
+│   └── core/
+│       ├── config.py            # pydantic-settings
+│       ├── http_client.py       # 共享 httpx 客户端
+│       └── retry.py             # 重试工具
+├── alembic/versions/            # 0001~0006(0001 建表 → 0006 漂移清理)
+├── frontend/src/pages/Matches.tsx  # 单页(预测面板 + 专家报告折叠区 + 游标分页)
+├── tests/                       # 核心 + agent 测试
 ├── docker-compose.yml           # api + postgres 两容器
 └── docs/                         # 本文档
 ```

docs/07-development.md

@@ -35,27 +35,34 @@ Profeto/
 ├── src/                        # 后端源码
 │   ├── api/
 │   │   ├── routes/             # FastAPI 路由
-│   │   │   ├── leagues.py      # 联赛/比赛查询
+│   │   │   ├── matches.py      # 联赛/比赛查询
 │   │   │   ├── predict.py      # 预测入口
 │   │   │   ├── ingest.py       # 数据采集
-│   │   │   └── eval.py         # 评估回填
+│   │   │   ├── eval.py         # 评估回填
+│   │   │   └── backtest.py     # 历史回测
+│   │   ├── deps.py             # 依赖:管理接口鉴权(X-API-Key)
 │   │   ├── schemas.py          # Pydantic 模型
 │   │   └── app.py              # FastAPI 工厂
 │   ├── db/
 │   │   ├── base.py             # SQLAlchemy async engine + session
-│   │   └── models.py           # 6 张表 ORM
+│   │   ├── models.py           # 6 张表 ORM
+│   │   ├── repositories.py     # 仓储层(查询封装)
+│   │   └── unit_of_work.py     # 事务边界
 │   ├── data/
 │   │   ├── bzzoiro.py          # bzzoiro 采集 + 入库
 │   │   ├── understat.py        # understat xG 回填
 │   │   ├── injuries.py         # 伤停采集
 │   │   ├── normalize.py        # 数据清洗契约
 │   │   ├── team_names.py       # 队名归一化映射
+│   │   ├── sources.py          # 数据源注册表
 │   │   └── config.py           # 联赛映射常量
 │   ├── llm/
 │   │   ├── provider.py         # LLM 提供商抽象(OpenAI-compatible)
 │   │   ├── context_builder.py  # 数据切片 + 拼接
 │   │   ├── predict.py          # 预测入口(单/多模式分派)
 │   │   ├── eval.py             # 评估统计
+│   │   ├── validation.py       # LLM 输出严格校验
+│   │   ├── backtest.py         # 回测执行
 │   │   ├── agents/
 │   │   │   ├── base.py         # AgentSpec + run_agent
 │   │   │   └── orchestrator.py # 多 agent 编排
@@ -69,12 +76,18 @@ Profeto/
 │   │           ├── h2h_v1.md
 │   │           └── aggregator_v1.md
 │   └── core/
-│       └── config.py           # pydantic-settings 配置
+│       ├── config.py           # pydantic-settings 配置
+│       ├── http_client.py      # 共享 httpx 客户端
+│       └── retry.py            # 重试工具
 ├── frontend/                   # React 单页前端
 ├── alembic/                    # 数据库迁移
 │   └── versions/
 │       ├── 0001_initial.py
-│       └── 0002_agent_outputs.py
+│       ├── 0002_agent_outputs.py
+│       ├── 0003_injuries.py
+│       ├── 0004_snapshot_and_constraints.py
+│       ├── 0005_prediction_status_and_stats_provenance.py
+│       └── 0006_schema_model_drift_cleanup.py
 ├── tests/                      # 测试
 │   ├── test_core.py            # 核心逻辑测试
 │   └── test_agents.py          # 多 agent 测试

frontend/postcss.config.js

@@ -1,5 +1,6 @@
 export default {
-  content: ['./index.html', './src/**/*.{ts,tsx}'],
-  theme: { extend: {} },
-  plugins: [],
+  plugins: {
+    tailwindcss: {},
+    autoprefixer: {},
+  },
 }

frontend/src/App.tsx

@@ -1,17 +1,54 @@
 import { ErrorBoundary } from './components/ErrorBoundary'
 import Matches from './pages/Matches'
 
+/** 报眉日期行:2026年9月15日 星期二 */
+function dateLine(): string {
+  return new Date().toLocaleDateString('zh-CN', {
+    year: 'numeric',
+    month: 'long',
+    day: 'numeric',
+    weekday: 'long',
+  })
+}
+
 export default function App() {
   return (
     <ErrorBoundary>
-      <div className="min-h-screen bg-gray-50">
-        <header className="bg-white border-b px-6 py-3 flex items-center justify-between">
-          <h1 className="text-xl font-bold text-blue-700">⚽ 先知 Profeto</h1>
-          <span className="text-sm text-gray-500">足球 LLM 预测服务</span>
+      <div className="min-h-screen bg-paper-50">
+        {/* ── 报头:粗线 + 居中刊名 + 报眉 ── */}
+        <header className="masthead-rule">
+          <div className="mx-auto max-w-5xl px-5 sm:px-8">
+            <div className="border-b border-ink-900 py-5 text-center sm:py-6">
+              <h1 className="font-serif text-4xl font-bold tracking-widest text-ink-900">
+                先知
+                <span className="ml-3 align-baseline font-serif text-base font-normal italic tracking-normal text-ink-500">
+                  Profeto
+                </span>
+              </h1>
+              <p className="mt-2 text-2xs tracking-[0.4em] text-ink-500">
+                足球比分预测 · 五路专家 · 终裁汇总
+              </p>
+            </div>
+            <div className="flex items-center justify-between border-b border-ink-200 py-2 text-2xs text-ink-500">
+              <span>{dateLine()}</span>
+              <span className="flex items-center gap-1.5">
+                <span className="inline-block h-1.5 w-1.5 bg-emerald-600" aria-hidden="true" />
+                服务运行中
+              </span>
+            </div>
+          </div>
         </header>
-        <main className="max-w-5xl mx-auto p-6">
+
+        <main className="mx-auto max-w-5xl px-5 py-6 sm:px-8 sm:py-8">
           <Matches />
         </main>
+
+        {/* ── 版底 ── */}
+        <footer className="mx-auto max-w-5xl px-5 pb-10 sm:px-8">
+          <div className="border-t border-ink-200 pt-3 text-center text-2xs leading-relaxed text-ink-400">
+            预测结果由大语言模型生成 · 仅供研究参考 · 不构成任何投注建议
+          </div>
+        </footer>
       </div>
     </ErrorBoundary>
   )

frontend/src/index.css

@@ -1,3 +1,102 @@
 @tailwind base;
 @tailwind components;
 @tailwind utilities;
+
+@layer base {
+  html {
+    -webkit-text-size-adjust: 100%;
+    text-rendering: optimizeLegibility;
+  }
+
+  body {
+    @apply bg-paper-50 text-ink-800 font-sans antialiased;
+    font-feature-settings: 'tnum' 1; /* 数字等宽:比分/百分比不跳动 */
+  }
+
+  ::selection {
+    @apply bg-press-wash text-press;
+  }
+
+  /* 统一焦点环:印报红细线,键盘可达 */
+  :focus-visible {
+    outline: 2px solid #9e1b1b;
+    outline-offset: 2px;
+  }
+
+  /* 细滚动条 */
+  ::-webkit-scrollbar {
+    width: 10px;
+    height: 10px;
+  }
+  ::-webkit-scrollbar-track {
+    background: transparent;
+  }
+  ::-webkit-scrollbar-thumb {
+    @apply rounded-full bg-ink-300;
+  }
+  ::-webkit-scrollbar-thumb:hover {
+    @apply bg-ink-400;
+  }
+
+  /* 尊重「减少动态效果」偏好 */
+  @media (prefers-reduced-motion: reduce) {
+    *,
+    *::before,
+    *::after {
+      animation-duration: 0.01ms !important;
+      animation-iteration-count: 1 !important;
+      transition-duration: 0.01ms !important;
+    }
+  }
+}
+
+@layer components {
+  /* ── 报头双线:粗线在上、细线在下 ── */
+  .masthead-rule {
+    border-top: 3px solid #17140f;
+  }
+
+  /* ── 按钮:方正边框式,悬停反白 ── */
+  .btn {
+    @apply inline-flex items-center justify-center gap-1.5 border border-ink-300 bg-transparent px-3 py-1.5
+           text-sm text-ink-700 transition-colors duration-150
+           hover:border-ink-900 hover:bg-ink-900 hover:text-paper-50
+           disabled:cursor-not-allowed disabled:opacity-40
+           disabled:hover:border-ink-300 disabled:hover:bg-transparent disabled:hover:text-ink-700;
+  }
+  .btn-sm {
+    @apply px-2.5 py-1 text-xs;
+  }
+  .btn-solid {
+    @apply border-ink-900 bg-ink-900 text-paper-50 hover:border-press hover:bg-press;
+  }
+
+  /* ── 表单控件:方正、无圆角 ── */
+  .field {
+    @apply border border-ink-300 bg-transparent px-2.5 py-1.5 text-sm text-ink-800
+           transition-colors hover:border-ink-400
+           focus:border-press focus:outline-none;
+  }
+
+  /* ── 版面切换文字标签(联赛/状态/模式) ── */
+  .tab {
+    @apply relative whitespace-nowrap px-0.5 py-1 text-sm text-ink-500 transition-colors hover:text-ink-900;
+  }
+  .tab-on {
+    @apply font-medium text-press;
+  }
+  .tab-on::after {
+    content: '';
+    @apply absolute inset-x-0 bottom-0 h-0.5 bg-press;
+  }
+
+  /* ── 小节标题:宋体加粗 + 墨色底线 ── */
+  .section-head {
+    @apply border-b border-ink-900 pb-1.5 font-serif text-sm font-bold text-ink-900;
+  }
+
+  /* ── 骨架占位:低调脉动,不用渐变扫光 ── */
+  .skeleton {
+    @apply animate-pulse rounded-none bg-ink-200;
+  }
+}

frontend/src/pages/Matches.tsx

@@ -1,4 +1,4 @@
-import { useCallback, useEffect, useState } from 'react'
+import { useCallback, useEffect, useRef, useState } from 'react'
 
 interface Match {
   id: number
@@ -65,54 +65,207 @@ const LEAGUES = [
   { code: 'F1', name: '法甲' },
 ]
 
+/** 汉字编号,给专家意见排版用 */
+const CN_NUM = ['一', '二', '三', '四', '五', '六', '七', '八']
+
+const STATUS_META: Record<string, { label: string; cls: string }> = {
+  finished: { label: '已完赛', cls: 'text-ink-400' },
+  scheduled: { label: '未开赛', cls: 'text-ink-600' },
+  live: { label: '进行中', cls: 'text-press font-medium' },
+}
+
+/** 1x2 → 中文标签 */
+const OUTCOME_LABEL: Record<string, string> = { '1': '主胜', X: '平局', '2': '客胜' }
+
+/** 置信度细线:0~1 数值的低调可视化 */
+function Meter({ value }: { value: number }) {
+  const pct = Math.max(0, Math.min(100, Math.round(value * 100)))
+  return (
+    <div className="h-px w-full bg-ink-200" role="presentation">
+      <div className="h-px bg-press transition-[width] duration-500" style={{ width: `${pct}%` }} />
+    </div>
+  )
+}
+
+/** home_edge(-1~1,正=利主队)的可视化:以中线为原点的双向细条 */
+function EdgeBar({ value }: { value: number }) {
+  const v = Math.max(-1, Math.min(1, value))
+  const half = Math.abs(v) * 50
+  return (
+    <div className="relative h-px w-full bg-ink-200" role="presentation">
+      <span className="absolute left-1/2 top-1/2 h-2 w-px -translate-x-1/2 -translate-y-1/2 bg-ink-400" />
+      <span
+        className={`absolute top-0 h-px transition-all duration-500 ${v >= 0 ? 'bg-press' : 'bg-ink-600'}`}
+        style={
+          v >= 0
+            ? { left: '50%', width: `${half}%` }
+            : { right: '50%', width: `${half}%` }
+        }
+      />
+    </div>
+  )
+}
+
+/** 骨架占位行:低调脉动灰块 */
+function SkeletonRows({ n = 4 }: { n?: number }) {
+  return (
+    <>
+      {Array.from({ length: n }).map((_, i) => (
+        <div key={i} className="flex items-center gap-4 border-b border-ink-200 px-1 py-3.5">
+          <div className="skeleton h-3 w-16" />
+          <div className="skeleton h-3 flex-1" />
+          <div className="skeleton h-3 w-10" />
+          <div className="skeleton h-3 flex-1" />
+          <div className="skeleton h-3 w-16" />
+        </div>
+      ))}
+    </>
+  )
+}
+
+function Spinner({ className = '' }: { className?: string }) {
+  return (
+    <svg
+      viewBox="0 0 20 20"
+      className={`h-3.5 w-3.5 animate-spin ${className}`}
+      fill="none"
+      aria-hidden="true"
+    >
+      <circle cx="10" cy="10" r="7.5" stroke="currentColor" strokeWidth="1.5" strokeOpacity="0.25" />
+      <path d="M17.5 10A7.5 7.5 0 0010 2.5" stroke="currentColor" strokeWidth="1.5" strokeLinecap="round" />
+    </svg>
+  )
+}
+
+/** 胜平负一行文字:选中的红字加方块标记,未选中的退灰 */
+function OutcomeLine({
+  pick,
+  confidence,
+}: {
+  pick: string | null
+  confidence: number | null
+}) {
+  const options = ['1', 'X', '2'] as const
+  return (
+    <div>
+      <div className="flex items-baseline justify-center gap-6 sm:gap-10">
+        {options.map(o => {
+          const on = pick === o
+          return (
+            <div key={o} className="flex flex-col items-center gap-1">
+              <span className={`flex items-center gap-1.5 text-sm ${on ? 'font-semibold text-press' : 'text-ink-400'}`}>
+                {on && <span className="inline-block h-2 w-2 bg-press" aria-hidden="true" />}
+                {OUTCOME_LABEL[o]}
+              </span>
+              {on && confidence !== null && (
+                <span className="text-2xs tabular-nums text-ink-500">
+                  置信 {Math.round(confidence * 100)}%
+                </span>
+              )}
+            </div>
+          )
+        })}
+      </div>
+      {pick && confidence !== null && (
+        <div className="mx-auto mt-3 max-w-xs">
+          <Meter value={confidence} />
+          <p className="mt-1 text-center text-2xs text-ink-400">主观置信度,非统计概率</p>
+        </div>
+      )}
+    </div>
+  )
+}
+
 export default function Matches() {
   const [league, setLeague] = useState('E0')
   const [status, setStatus] = useState('scheduled')
   const [matches, setMatches] = useState<Match[]>([])
+  const [nextCursor, setNextCursor] = useState<string | null>(null)
+  const [loadingMore, setLoadingMore] = useState(false)
   const [loading, setLoading] = useState(false)
   const [predictingId, setPredictingId] = useState<number | null>(null)
   const [prediction, setPrediction] = useState<Prediction | null>(null)
+  const [predictionFor, setPredictionFor] = useState<Match | null>(null)
   const [error, setError] = useState<string | null>(null)
   const [mode, setMode] = useState<'single' | 'multi'>('multi')
 
+  // 请求竞态防护:切换联赛/状态很快时,先发的慢请求可能后返回,
+  // 把旧结果覆盖到新筛选上。用递增序号只认最后一次请求的响应。
+  const loadSeq = useRef(0)
+  const predictSeq = useRef(0)
+
   const load = useCallback(async () => {
+    const seq = ++loadSeq.current
     setLoading(true)
+    // 切换筛选时作废进行中的「加载更多」,避免其标志位卡住
+    setLoadingMore(false)
     setError(null)
     try {
       const params = new URLSearchParams({ league, status, limit: '50' })
       const res = await fetch(`/api/v1/matches?${params}`)
+      if (seq !== loadSeq.current) return // 已有更新的请求,丢弃本次结果
       if (!res.ok) throw new Error(`HTTP ${res.status}`)
       const data = await res.json()
+      if (seq !== loadSeq.current) return
       setMatches(data.items)
+      setNextCursor(data.next_cursor ?? null)
     } catch (e) {
+      if (seq !== loadSeq.current) return
       setError(e instanceof Error ? e.message : String(e))
     } finally {
-      setLoading(false)
+      if (seq === loadSeq.current) setLoading(false)
     }
   }, [league, status])
 
+  // 加载下一页(游标分页)
+  const loadMore = async () => {
+    if (!nextCursor || loadingMore) return
+    const seq = loadSeq.current // 不做自增:切换筛选会自增,这里只跟随当前序列
+    setLoadingMore(true)
+    try {
+      const params = new URLSearchParams({ league, status, limit: '50', cursor: nextCursor })
+      const res = await fetch(`/api/v1/matches?${params}`)
+      if (seq !== loadSeq.current) return
+      if (!res.ok) throw new Error(`HTTP ${res.status}`)
+      const data = await res.json()
+      if (seq !== loadSeq.current) return
+      setMatches(prev => [...prev, ...data.items])
+      setNextCursor(data.next_cursor ?? null)
+    } catch (e) {
+      if (seq !== loadSeq.current) return
+      setError(e instanceof Error ? e.message : String(e))
+    } finally {
+      if (seq === loadSeq.current) setLoadingMore(false)
+    }
+  }
+
   useEffect(() => { load() }, [load])
 
-  const predict = async (matchId: number) => {
-    setPredictingId(matchId)
+  const predict = async (m: Match) => {
+    const seq = ++predictSeq.current
+    setPredictingId(m.id)
     setError(null)
     setPrediction(null)
+    setPredictionFor(m)
     try {
       const res = await fetch('/api/v1/predict', {
         method: 'POST',
         headers: { 'Content-Type': 'application/json' },
-        body: JSON.stringify({ match_id: matchId, mode }),
+        body: JSON.stringify({ match_id: m.id, mode }),
       })
+      if (seq !== predictSeq.current) return
       if (!res.ok) {
         const t = await res.text()
         throw new Error(`HTTP ${res.status}: ${t}`)
       }
       const data = await res.json()
+      if (seq !== predictSeq.current) return
       setPrediction(data)
     } catch (e) {
+      if (seq !== predictSeq.current) return
       setError(e instanceof Error ? e.message : String(e))
     } finally {
-      setPredictingId(null)
+      if (seq === predictSeq.current) setPredictingId(null)
     }
   }
 
@@ -121,195 +274,410 @@ export default function Matches() {
     return d.toLocaleString('zh-CN', { month: '2-digit', day: '2-digit', hour: '2-digit', minute: '2-digit' })
   }
 
+  const leagueName = LEAGUES.find(l => l.code === league)?.name ?? league
+
+  /** 状态/模式一组的文字切换 */
+  const Switch = ({ value, onChange, items }: {
+    value: string
+    onChange: (v: string) => void
+    items: { v: string; label: string; title?: string }[]
+  }) => (
+    <span className="inline-flex items-center gap-2.5">
+      {items.map((it, i) => (
+        <span key={it.v} className="inline-flex items-center gap-2.5">
+          {i > 0 && <span className="text-ink-300" aria-hidden="true">/</span>}
+          <button
+            onClick={() => onChange(it.v)}
+            title={it.title}
+            className={`relative tab ${value === it.v ? 'tab-on' : ''} text-xs`}
+          >
+            {it.label}
+          </button>
+        </span>
+      ))}
+    </span>
+  )
+
   return (
-    <div className="space-y-4">
-      {/* 筛选 */}
-      <div className="flex gap-3 items-center flex-wrap">
-        <select value={league} onChange={e => setLeague(e.target.value)}
-          className="border rounded px-3 py-1.5 text-sm">
-          {LEAGUES.map(l => <option key={l.code} value={l.code}>{l.name}</option>)}
-        </select>
-        <select value={status} onChange={e => setStatus(e.target.value)}
-          className="border rounded px-3 py-1.5 text-sm">
-          <option value="scheduled">未开赛</option>
-          <option value="finished">已完赛</option>
-          <option value="">全部</option>
-        </select>
-        <div className="flex items-center gap-2 text-sm">
-          <span className="text-gray-500">模式:</span>
+    <div className="space-y-5">
+      {/* ── 联赛版面切换 ── */}
+      <nav className="flex items-center gap-6 overflow-x-auto border-b border-ink-900" aria-label="联赛">
+        {LEAGUES.map(l => (
           <button
-            onClick={() => setMode('single')}
-            className={`px-2 py-1 rounded text-xs ${mode === 'single' ? 'bg-blue-600 text-white' : 'bg-gray-100 text-gray-600'}`}
-          >单次</button>
-          <button
-            onClick={() => setMode('multi')}
-            className={`px-2 py-1 rounded text-xs ${mode === 'multi' ? 'bg-blue-600 text-white' : 'bg-gray-100 text-gray-600'}`}
-          >多 Agent</button>
-        </div>
-        <button onClick={load} disabled={loading}
-          className="bg-blue-600 text-white text-sm px-4 py-1.5 rounded hover:bg-blue-700 disabled:opacity-50">
-          {loading ? '加载中...' : '刷新'}
-        </button>
-        <span className="text-sm text-gray-500">共 {matches.length} 场</span>
+            key={l.code}
+            onClick={() => setLeague(l.code)}
+            className={`relative tab ${league === l.code ? 'tab-on' : ''} font-serif`}
+          >
+            {l.name}
+          </button>
+        ))}
+      </nav>
+
+      {/* ── 第二行:状态 / 模式 / 计数 / 刷新 ── */}
+      <div className="flex flex-wrap items-center gap-x-5 gap-y-2 text-xs text-ink-500">
+        <span className="inline-flex items-center gap-2.5">
+          <span className="text-2xs text-ink-400">状态</span>
+          <Switch
+            value={status}
+            onChange={setStatus}
+            items={[
+              { v: 'scheduled', label: '未开赛' },
+              { v: 'finished', label: '已完赛' },
+              { v: '', label: '全部' },
+            ]}
+          />
+        </span>
+
+        <span className="inline-flex items-center gap-2.5">
+          <span className="text-2xs text-ink-400">模式</span>
+          <Switch
+            value={mode}
+            onChange={v => setMode(v as 'single' | 'multi')}
+            items={[
+              { v: 'single', label: '单次', title: '单次调用,快但只有一个模型看全部数据' },
+              { v: 'multi', label: '多专家', title: '5 个专家并行分析后由终裁汇总,质量更高' },
+            ]}
+          />
+        </span>
+
+        <span className="ml-auto inline-flex items-center gap-3">
+          <span className="tabular-nums">共 {matches.length} 场</span>
+          <button onClick={load} disabled={loading} className="btn btn-sm">
+            {loading ? (<><Spinner /> 获取中</>) : '刷新'}
+          </button>
+        </span>
       </div>
 
+      {/* ── 错误提示 ── */}
       {error && (
-        <div className="bg-red-50 border border-red-200 text-red-700 px-4 py-2 rounded text-sm flex items-center justify-between">
-          <span>❌ {error}</span>
-          <button onClick={() => setError(null)} className="text-red-500 hover:text-red-700 text-xs">✕</button>
+        <div className="flex items-start justify-between gap-3 border border-press bg-press-wash px-4 py-3">
+          <div>
+            <p className="text-sm font-medium text-press">请求失败</p>
+            <p className="mt-0.5 text-xs text-ink-600">{error}</p>
+          </div>
+          <button onClick={() => setError(null)} className="text-ink-400 transition-colors hover:text-ink-900" aria-label="关闭">
+            <svg viewBox="0 0 20 20" className="h-4 w-4" fill="currentColor" aria-hidden="true">
+              <path d="M6.3 5.3a1 1 0 011.4 0L10 7.6l2.3-2.3a1 1 0 111.4 1.4L11.4 9l2.3 2.3a1 1 0 01-1.4 1.4L10 10.4l-2.3 2.3a1 1 0 01-1.4-1.4L8.6 9 6.3 6.7a1 1 0 010-1.4z" />
+            </svg>
+          </button>
         </div>
       )}
 
-      {/* 比赛表 */}
-      <div className="bg-white rounded border overflow-hidden">
-        <table className="w-full text-sm">
-          <thead className="bg-gray-100 text-gray-600">
-            <tr>
-              <th className="text-left px-4 py-2">日期</th>
-              <th className="text-left px-4 py-2">主队</th>
-              <th className="text-left px-4 py-2">客队</th>
-              <th className="text-center px-4 py-2">比分</th>
-              <th className="text-center px-4 py-2">状态</th>
-              <th className="text-center px-4 py-2">操作</th>
-            </tr>
-          </thead>
-          <tbody>
-            {loading && (
-              <tr><td colSpan={6} className="text-center text-gray-400 py-8">
-                <span className="inline-block animate-spin mr-2">⏳</span>加载中...
-              </td></tr>
-            )}
-            {!loading && matches.length === 0 && (
-              <tr><td colSpan={6} className="text-center text-gray-400 py-8">暂无数据,请先采集</td></tr>
-            )}
-            {matches.map(m => (
-              <tr key={m.id} className="border-t hover:bg-gray-50">
-                <td className="px-4 py-2 text-gray-600">{fmtDate(m.match_date)}</td>
-                <td className="px-4 py-2 font-medium">{m.home_team_zh || m.home_team}</td>
-                <td className="px-4 py-2 font-medium">{m.away_team_zh || m.away_team}</td>
-                <td className="px-4 py-2 text-center">
-                  {m.home_goals !== null ? `${m.home_goals} - ${m.away_goals}` : '-'}
-                </td>
-                <td className="px-4 py-2 text-center">
-                  <span className={`text-xs px-2 py-0.5 rounded ${
-                    m.match_status === 'finished' ? 'bg-green-100 text-green-700' :
-                    m.match_status === 'scheduled' ? 'bg-blue-100 text-blue-700' : 'bg-gray-100'
-                  }`}>
-                    {m.match_status === 'finished' ? '完赛' : m.match_status === 'scheduled' ? '未开赛' : m.match_status}
-                  </span>
-                </td>
-                <td className="px-4 py-2 text-center">
-                  <button onClick={() => predict(m.id)}
-                    disabled={predictingId === m.id}
-                    className="text-blue-600 hover:underline text-xs disabled:opacity-50">
-                    {predictingId === m.id ? '预测中...' : 'LLM 预测'}
+      {/* ── 赛程栏:表格化,行间细线 ── */}
+      <section aria-label="赛程">
+        {loading && <SkeletonRows n={4} />}
+
+        {!loading && matches.length === 0 && (
+          <div className="border-y border-ink-200 py-14 text-center">
+            <p className="font-serif text-sm text-ink-600">本版暂无赛程</p>
+            <p className="mt-1.5 text-xs text-ink-400">请先通过采集接口导入 {leagueName} 的比赛数据</p>
+          </div>
+        )}
+
+        {!loading && matches.map(m => {
+          const st = STATUS_META[m.match_status] ?? { label: m.match_status, cls: 'text-ink-400' }
+          const homeName = m.home_team_zh || m.home_team
+          const awayName = m.away_team_zh || m.away_team
+          const busy = predictingId === m.id
+          const active = predictionFor?.id === m.id
+
+          return (
+            <div
+              key={m.id}
+              className={`border-b border-ink-200 px-1 py-3 transition-colors hover:bg-paper-100 ${
+                active ? 'bg-press-wash/50' : ''
+              }`}
+            >
+              <div className="flex flex-col gap-2 sm:grid sm:grid-cols-[88px_minmax(0,1fr)_64px_minmax(0,1fr)_56px_auto] sm:items-center sm:gap-x-3 sm:gap-y-0">
+                {/* 日期 + 状态:移动端同行,桌面端日期单独归列 */}
+                <div className="flex items-center justify-between sm:contents">
+                  <span className="text-2xs tabular-nums text-ink-400">{fmtDate(m.match_date)}</span>
+                  <span className={`text-2xs sm:hidden ${st.cls}`}>{st.label}</span>
+                </div>
+
+                {/* 对阵:移动端主队/比分/客队同一行,桌面端 sm:contents 拆回 grid 列 */}
+                <div className="flex items-center gap-2 sm:contents">
+                  {/* 主队(右对齐) */}
+                  <div className="flex min-w-0 flex-1 items-center justify-end">
+                    <span className="truncate text-sm font-medium text-ink-900">{homeName}</span>
+                  </div>
+
+                  {/* 比分 / VS */}
+                  <div className="flex w-14 flex-shrink-0 flex-col items-center sm:w-auto">
+                    {m.home_goals !== null && m.away_goals !== null ? (
+                      <span className="font-serif text-base font-bold tabular-nums text-ink-900">
+                        {m.home_goals}<span className="mx-0.5 font-normal text-ink-300">:</span>{m.away_goals}
+                      </span>
+                    ) : (
+                      <span className="text-2xs tracking-widest text-ink-400">VS</span>
+                    )}
+                    {m.home_xg !== null && m.away_xg !== null && (
+                      <span className="text-2xs tabular-nums text-ink-400">
+                        xG {m.home_xg.toFixed(1)}-{m.away_xg.toFixed(1)}
+                      </span>
+                    )}
+                  </div>
+
+                  {/* 客队(左对齐) */}
+                  <div className="flex min-w-0 flex-1 items-center">
+                    <span className="truncate text-sm font-medium text-ink-900">{awayName}</span>
+                  </div>
+                </div>
+
+                {/* 状态列(桌面) */}
+                <span className={`hidden text-right text-2xs sm:block ${st.cls}`}>{st.label}</span>
+
+                {/* 预测按钮 */}
+                <div className="flex justify-end">
+                  <button
+                    onClick={() => predict(m)}
+                    disabled={busy}
+                    className="btn btn-sm w-[76px]"
+                    title={`以${mode === 'multi' ? '多专家' : '单次'}模式预测这场`}
+                  >
+                    {busy ? (<><Spinner /> 预测中</>) : '预测'}
                   </button>
-                </td>
-              </tr>
-            ))}
-          </tbody>
-        </table>
-      </div>
-
-      {/* 预测中指示 */}
-      {predictingId && (
-        <div className="bg-blue-50 border border-blue-200 text-blue-700 px-4 py-3 rounded text-sm flex items-center gap-2">
-          <span className="inline-block animate-spin">⏳</span>
-          LLM 预测中,请稍候...
-        </div>
-      )}
-
-      {/* 预测结果 */}
-      {prediction && (
-        <div className="bg-white rounded border p-5 space-y-3">
-          <h3 className="font-bold text-lg">🤖 LLM 预测结果</h3>
-          <div className="grid grid-cols-2 md:grid-cols-4 gap-4 text-sm">
-            <div className="bg-blue-50 rounded p-3">
-              <div className="text-gray-500 text-xs">主进球</div>
-              <div className="text-xl font-bold">{prediction.pred_home_goals ?? '-'}</div>
-            </div>
-            <div className="bg-blue-50 rounded p-3">
-              <div className="text-gray-500 text-xs">客进球</div>
-              <div className="text-xl font-bold">{prediction.pred_away_goals ?? '-'}</div>
-            </div>
-            <div className="bg-amber-50 rounded p-3">
-              <div className="text-gray-500 text-xs">胜平负</div>
-              <div className="text-xl font-bold">{prediction.pred_1x2 ?? '-'}</div>
-            </div>
-            <div className="bg-green-50 rounded p-3">
-              <div className="text-gray-500 text-xs">置信度</div>
-              <div className="text-xl font-bold">
-                {prediction.subjective_confidence !== null ? `${(prediction.subjective_confidence * 100).toFixed(0)}%` : '-'}
+                </div>
               </div>
             </div>
-          </div>
-          <div className="text-xs text-gray-400">
-            {prediction.provider} / {prediction.model} · 耗时 {prediction.latency_ms}ms
-            {prediction.mode === 'multi' && ' · 多 Agent 模式'}
-          </div>
+          )
+        })}
 
-          {/* 各专家 agent 报告 */}
-          {prediction.agent_outputs && prediction.agent_outputs.length > 0 && (
-            <div className="space-y-2">
-              <div className="text-sm font-medium text-gray-700">专家 Agent 报告</div>
-              {prediction.agent_outputs.map((r) => (
-                <details key={r.agent} className="bg-white border rounded">
-                  <summary className="cursor-pointer px-3 py-2 text-sm flex items-center justify-between">
-                    <span className="font-medium">
-                      {AGENT_LABELS[r.agent] || r.agent}
-                      {r.status !== 'ok' && (
-                        <span className={`ml-2 text-xs px-1.5 py-0.5 rounded ${
-                          r.status === 'no_data' ? 'bg-gray-100 text-gray-500' : 'bg-red-100 text-red-600'
-                        }`}>
-                          {r.status === 'no_data' ? '无数据' : '失败'}
-                        </span>
-                      )}
-                    </span>
-                    <span className="flex gap-3 text-xs text-gray-500">
-                      {r.home_edge !== null && (
-                        <span className={r.home_edge > 0 ? 'text-blue-600' : r.home_edge < 0 ? 'text-amber-600' : ''}>
-                          主队优势 {r.home_edge > 0 ? '+' : ''}{r.home_edge.toFixed(2)}
-                        </span>
-                      )}
-                      {r.subjective_confidence !== null && <span>信心 {(r.subjective_confidence * 100).toFixed(0)}%</span>}
-                      {r.probable_score && <span>比分 {r.probable_score}</span>}
-                    </span>
-                  </summary>
-                  <div className="px-3 pb-3 pt-1 space-y-2 text-sm">
-                    {r.analysis && <p className="text-gray-700">{r.analysis}</p>}
-                    {r.key_evidence.length > 0 && (
-                      <ul className="text-xs text-gray-500 list-disc pl-4">
-                        {r.key_evidence.map((e, i) => <li key={i}>{e}</li>)}
-                      </ul>
-                    )}
-                    {r.exp_home_goals !== null && r.exp_away_goals !== null && (
-                      <div className="text-xs text-gray-500">
-                        进球期望: {r.exp_home_goals.toFixed(1)} - {r.exp_away_goals.toFixed(1)}
-                      </div>
-                    )}
-                    <div className="text-xs text-gray-400">
-                      数据充分度 {r.data_sufficiency} · {r.model} · {r.latency_ms}ms
-                    </div>
-                  </div>
-                </details>
-              ))}
-            </div>
-          )}
+        {!loading && nextCursor && (
+          <div className="flex justify-center pt-4">
+            <button onClick={loadMore} disabled={loadingMore} className="btn btn-sm">
+              {loadingMore ? (<><Spinner /> 获取中</>) : '载入更多'}
+            </button>
+          </div>
+        )}
+      </section>
 
-          {prediction.reasoning && (
-            <div className="bg-gray-50 rounded p-3">
-              <div className="text-xs text-gray-500 mb-1">推理过程</div>
-              <div className="text-sm whitespace-pre-wrap">{prediction.reasoning}</div>
+      {/* ── 预测中占位 ── */}
+      {predictingId && !prediction && (
+        <div className="border border-ink-900">
+          <div className="flex items-center gap-2 border-b border-ink-900 bg-paper-100 px-4 py-2.5">
+            <Spinner className="text-press" />
+            <span className="text-sm font-medium text-ink-800">正在生成预测</span>
+            <span className="text-2xs text-ink-500">
+              {mode === 'multi' ? '五路专家并行分析后终裁,约需 20-60 秒' : '单次调用,约需 5-15 秒'}
+            </span>
+          </div>
+          <div className="space-y-4 px-4 py-6">
+            <div className="flex items-center justify-center gap-6">
+              <div className="skeleton h-4 w-20" />
+              <div className="skeleton h-10 w-24" />
+              <div className="skeleton h-4 w-20" />
             </div>
-          )}
-          <details className="text-xs">
-            <summary className="cursor-pointer text-gray-500 hover:text-gray-700">查看完整上下文</summary>
-            <pre className="mt-2 bg-gray-900 text-green-300 p-3 rounded overflow-x-auto text-xs">
-              {prediction.context}
-            </pre>
-          </details>
+            <div className="skeleton mx-auto h-px w-64" />
+            <div className="skeleton h-16 w-full" />
+          </div>
         </div>
       )}
+
+      {/* ── 预测版 ── */}
+      {prediction && predictionFor && (
+        <PredictionPanel prediction={prediction} match={predictionFor} mode={mode} />
+      )}
     </div>
   )
 }
+
+function PredictionPanel({
+  prediction,
+  match,
+  mode,
+}: {
+  prediction: Prediction
+  match: Match
+  mode: 'single' | 'multi'
+}) {
+  const homeName = match.home_team_zh || match.home_team
+  const awayName = match.away_team_zh || match.away_team
+  const okReports = (prediction.agent_outputs ?? []).filter(r => r.status === 'ok')
+
+  return (
+    <article className="border border-ink-900 bg-paper-50">
+      {/* 版头 */}
+      <div className="flex flex-wrap items-baseline justify-between gap-2 border-b border-ink-900 bg-paper-100 px-4 py-2.5 sm:px-5">
+        <h3 className="font-serif text-sm font-bold text-ink-900">
+          预测版 · {homeName} 对 {awayName}
+        </h3>
+        <span className="text-2xs tabular-nums text-ink-500">
+          {prediction.provider} / {prediction.model}
+          {prediction.latency_ms !== null && ` · ${(prediction.latency_ms / 1000).toFixed(1)}s`}
+        </span>
+      </div>
+
+      <div className="space-y-7 px-4 py-6 sm:px-5">
+        {/* ── 预测比分:版面核心,大号宋体 ── */}
+        <div className="text-center">
+          <p className="font-serif text-5xl font-bold tabular-nums leading-none text-ink-900 sm:text-6xl">
+            {prediction.pred_home_goals ?? '-'}
+            <span className="mx-3 font-normal text-ink-300">:</span>
+            {prediction.pred_away_goals ?? '-'}
+          </p>
+          <p className="mt-3 text-2xs tracking-[0.5em] text-ink-400">预测比分</p>
+        </div>
+
+        {/* ── 胜平负 ── */}
+        <div className="border-y border-ink-200 py-4">
+          <OutcomeLine pick={prediction.pred_1x2} confidence={prediction.subjective_confidence} />
+        </div>
+
+        {/* ── 元信息一行 ── */}
+        <p className="text-center text-2xs text-ink-500">
+          {mode === 'multi' ? `多专家模式 · ${okReports.length}/${prediction.agent_outputs?.length ?? 0} 路有效` : '单次模式'}
+          {prediction.prompt_version && ` · prompt ${prediction.prompt_version}`}
+          {prediction.latency_ms !== null && ` · 终裁耗时 ${(prediction.latency_ms / 1000).toFixed(1)} 秒`}
+        </p>
+
+        {/* ── 专家意见 ── */}
+        {prediction.agent_outputs && prediction.agent_outputs.length > 0 && (
+          <section>
+            <div className="section-head flex flex-wrap items-baseline justify-between gap-1">
+              <span>五路专家意见</span>
+              {prediction.agent_weights && (
+                <span className="font-sans text-2xs font-normal text-ink-500">
+                  终裁权重:{Object.entries(prediction.agent_weights)
+                    .sort((a, b) => b[1] - a[1])
+                    .map(([k, v]) => `${AGENT_LABELS[k] ?? k} ${Math.round(v * 100)}%`)
+                    .join(' / ')}
+                </span>
+              )}
+            </div>
+
+            <div>
+              {prediction.agent_outputs.map((r, i) => (
+                <AgentCard key={r.agent} report={r} no={CN_NUM[i] ?? String(i + 1)} />
+              ))}
+            </div>
+          </section>
+        )}
+
+        {/* ── 终裁意见:引文式,红竖线 ── */}
+        {prediction.reasoning && (
+          <section>
+            <h4 className="section-head mb-3">终裁意见</h4>
+            <blockquote className="border-l-2 border-press pl-4">
+              <p className="whitespace-pre-wrap font-serif text-sm leading-loose text-ink-700">
+                {prediction.reasoning}
+              </p>
+            </blockquote>
+          </section>
+        )}
+
+        {/* ── 原始上下文 ── */}
+        <details className="group">
+          <summary className="flex cursor-pointer list-none items-center gap-1.5 text-xs text-ink-500 transition-colors hover:text-ink-800">
+            <svg viewBox="0 0 20 20" className="h-3 w-3 transition-transform group-open:rotate-90" fill="currentColor" aria-hidden="true">
+              <path d="M7.3 5.3a1 1 0 011.4 0l4 4a1 1 0 010 1.4l-4 4a1 1 0 01-1.4-1.4L10.6 10 7.3 6.7a1 1 0 010-1.4z" />
+            </svg>
+            查看喂给模型的完整数据切片
+          </summary>
+          <pre className="mt-2 max-h-80 overflow-auto border border-ink-200 bg-paper-100 p-3 font-mono text-2xs leading-relaxed text-ink-600">
+            {prediction.context}
+          </pre>
+        </details>
+      </div>
+    </article>
+  )
+}
+
+const STATUS_BADGE: Record<string, { label: string; cls: string }> = {
+  ok: { label: '正常', cls: 'text-ink-500' },
+  no_data: { label: '无数据', cls: 'text-ink-400' },
+  error: { label: '调用失败', cls: 'text-press' },
+  parse_error: { label: '解析失败', cls: 'text-press' },
+}
+
+const SUFFICIENCY_LABEL: Record<string, string> = {
+  high: '充分',
+  medium: '一般',
+  low: '偏少',
+  none: '无',
+}
+
+/** 单路专家意见:汉字编号 + 细线行 */
+function AgentCard({ report: r, no }: { report: AgentReport; no: string }) {
+  const badge = STATUS_BADGE[r.status] ?? { label: r.status, cls: 'text-ink-400' }
+  const inactive = r.status !== 'ok'
+
+  return (
+    <details className="group border-b border-ink-200">
+      <summary className="flex cursor-pointer list-none items-baseline gap-2.5 px-1 py-3">
+        <span className="font-serif text-sm text-ink-400">{no}</span>
+        <span className="text-sm font-medium text-ink-900">{AGENT_LABELS[r.agent] ?? r.agent}</span>
+        <span className={`text-2xs ${badge.cls}`}>{badge.label}</span>
+
+        <span className="ml-auto flex items-baseline gap-3 text-2xs tabular-nums text-ink-500">
+          {r.status === 'ok' && r.subjective_confidence !== null && (
+            <span>信心 {Math.round(r.subjective_confidence * 100)}%</span>
+          )}
+          {r.status === 'ok' && r.probable_score && (
+            <span className="font-serif font-bold text-ink-800">{r.probable_score}</span>
+          )}
+          <svg viewBox="0 0 20 20" className="h-3 w-3 self-center text-ink-300 transition-transform group-open:rotate-90" fill="currentColor" aria-hidden="true">
+            <path d="M7.3 5.3a1 1 0 011.4 0l4 4a1 1 0 010 1.4l-4 4a1 1 0 01-1.4-1.4L10.6 10 7.3 6.7a1 1 0 010-1.4z" />
+          </svg>
+        </span>
+      </summary>
+
+      <div className="space-y-3 px-1 pb-4 pl-7">
+        {/* 无数据 / 失败时给出明确说明,避免用户以为是空白 bug */}
+        {inactive && (
+          <p className="text-xs leading-relaxed text-ink-500">
+            {r.status === 'no_data' && '该维度没有可用数据,已跳过 LLM 分析以节省额度(不影响其他专家)。'}
+            {r.status === 'error' && '该专家调用失败,本次结论未纳入其视角(fail-open 设计,不阻断整体预测)。'}
+            {r.status === 'parse_error' && '模型输出未通过格式校验,该报告已丢弃。'}
+          </p>
+        )}
+
+        {!inactive && r.home_edge !== null && (
+          <div>
+            <div className="mb-1.5 flex items-baseline justify-between text-2xs">
+              <span className="text-ink-500">主队优势</span>
+              <span className={`font-semibold tabular-nums ${r.home_edge > 0 ? 'text-press' : r.home_edge < 0 ? 'text-ink-700' : 'text-ink-500'}`}>
+                {r.home_edge > 0 ? '+' : ''}{r.home_edge.toFixed(2)}
+              </span>
+            </div>
+            <EdgeBar value={r.home_edge} />
+            <div className="mt-1 flex justify-between text-2xs text-ink-400">
+              <span>利客队</span>
+              <span>利主队</span>
+            </div>
+          </div>
+        )}
+
+        {r.analysis && (
+          <p className="font-serif text-sm leading-loose text-ink-700">{r.analysis}</p>
+        )}
+
+        {r.key_evidence.length > 0 && (
+          <ul className="space-y-1.5">
+            {r.key_evidence.map((e, i) => (
+              <li key={i} className="flex gap-2 text-xs leading-relaxed text-ink-600">
+                <span className="flex-shrink-0 text-ink-300" aria-hidden="true">—</span>
+                <span>{e}</span>
+              </li>
+            ))}
+          </ul>
+        )}
+
+        {r.exp_home_goals !== null && r.exp_away_goals !== null && (
+          <p className="text-xs text-ink-500">
+            进球期望 <span className="font-serif font-bold tabular-nums text-ink-900">{r.exp_home_goals.toFixed(1)} - {r.exp_away_goals.toFixed(1)}</span>
+          </p>
+        )}
+
+        {!inactive && (
+          <p className="border-t border-ink-100 pt-2.5 text-2xs text-ink-400">
+            数据充分度 {SUFFICIENCY_LABEL[r.data_sufficiency] ?? r.data_sufficiency}
+            <span className="mx-2 text-ink-200">|</span>
+            <span className="font-mono">{r.model}</span>
+            {r.latency_ms !== null && <span className="ml-2 tabular-nums">{r.latency_ms}ms</span>}
+          </p>
+        )}
+      </div>
+    </details>
+  )
+}

frontend/tailwind.config.js

@@ -1,6 +1,61 @@
 /** @type {import('tailwindcss').Config} */
 export default {
   content: ['./index.html', './src/**/*.{ts,tsx}'],
-  theme: { extend: {} },
+  theme: {
+    extend: {
+      colors: {
+        // 纸白:微暖底色,像新闻纸而不是纯白画布
+        paper: {
+          50: '#FDFCF8',
+          100: '#F7F4EC',
+          200: '#EDE9DE',
+          300: '#DDD7C7',
+        },
+        // 墨色:暖黑灰阶,替代冷调 slate
+        ink: {
+          50: '#FAF9F7',
+          100: '#F0EEE9',
+          200: '#E2DFD7',
+          300: '#C9C4B8',
+          400: '#9C9587',
+          500: '#6E675B',
+          600: '#524C42',
+          700: '#3B362E',
+          800: '#282420',
+          900: '#17140F',
+        },
+        // 印报红:全站唯一强调色,克制使用
+        press: {
+          DEFAULT: '#9E1B1B',
+          dark: '#7C1414',
+          wash: '#F7E9E4',
+        },
+      },
+      fontFamily: {
+        // 标题与比分:宋体血统,报纸版面的骨架
+        serif: [
+          'Georgia',
+          '"Times New Roman"',
+          '"Songti SC"',
+          'STSong',
+          'SimSun',
+          '"Noto Serif CJK SC"',
+          'serif',
+        ],
+        // 正文:中文黑体栈
+        sans: [
+          '"PingFang SC"',
+          '"Hiragino Sans GB"',
+          '"Microsoft YaHei"',
+          '"Noto Sans CJK SC"',
+          'sans-serif',
+        ],
+        mono: ['ui-monospace', 'SFMono-Regular', 'Menlo', 'Consolas', 'monospace'],
+      },
+      fontSize: {
+        '2xs': ['11px', { lineHeight: '16px' }],
+      },
+    },
+  },
   plugins: [],
 }

src/api/app.py

@@ -16,6 +16,9 @@ async def lifespan(app: FastAPI) -> AsyncIterator[None]:
     from src.db.base import init_db
     from src.core.http_client import close_client
     await init_db()  # 验证连接,不建表
+    # P2-5: 启动时执行一次 ingest failure 重试清理
+    from src.data.retry_worker import run_retry_worker
+    await run_retry_worker()
     yield
     await close_client()
 
@@ -29,12 +32,14 @@ def create_app() -> FastAPI:
     )
 
     origins = [o.strip() for o in settings.CORS_ORIGINS.split(",") if o.strip()]
+    methods = [m.strip() for m in settings.CORS_METHODS.split(",") if m.strip()]
+    headers = [h.strip() for h in settings.CORS_HEADERS.split(",") if h.strip()]
     app.add_middleware(
         CORSMiddleware,
         allow_origins=origins,
         allow_credentials=True,
-        allow_methods=["*"],
-        allow_headers=["*"],
+        allow_methods=methods,
+        allow_headers=headers,
     )
 
     from src.api.routes.matches import router as matches_router

src/api/deps.py

@@ -0,0 +1,37 @@
+"""API 依赖:鉴权等横切关注点。
+
+审查报告 P2-7:ingest / backtest / settle 这类「写入型或高成本」接口此前
+完全无鉴权 —— 任何能访问到服务的人都可触发采集、或直接烧掉 LLM 额度。
+
+策略(渐进式,不破坏本地开发):
+  - `ADMIN_API_KEY` 未配置 → 直接放行,并打一次 warning。
+    这样本地 `docker compose up` 无需额外配置即可用。
+  - 已配置 → 必须带匹配的 `X-API-Key` 请求头,否则 401。
+"""
+from __future__ import annotations
+
+import logging
+import secrets
+
+from fastapi import Header, HTTPException
+
+from src.core.config import settings
+
+logger = logging.getLogger(__name__)
+
+
+async def require_admin_key(x_api_key: str | None = Header(None, alias="X-API-Key")) -> None:
+    """保护「写入型 / 高成本」接口的依赖。
+
+    用法: `@router.post("/ingest/bzzoiro", dependencies=[Depends(require_admin_key)])`
+    """
+    expected = settings.ADMIN_API_KEY
+    if not expected:
+        logger.warning(
+            "ADMIN_API_KEY 未设置,采集/回测接口当前【无鉴权】。"
+            "生产环境请设置该环境变量。"
+        )
+        return
+
+    if not x_api_key or not secrets.compare_digest(x_api_key, expected):
+        raise HTTPException(status_code=401, detail="无效或缺失的 X-API-Key")

src/api/routes/backtest.py

@@ -3,9 +3,10 @@ from __future__ import annotations
 
 import logging
 
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter, Depends, HTTPException
 from pydantic import BaseModel, Field
 
+from src.api.deps import require_admin_key
 from src.llm.backtest import run_backtest
 
 logger = logging.getLogger(__name__)
@@ -22,10 +23,13 @@ class BacktestRequest(BaseModel):
     model: str | None = Field(None, description="指定模型 (空=默认)")
 
 
-@router.post("/backtest")
+@router.post("/backtest", dependencies=[Depends(require_admin_key)])
 async def backtest(req: BacktestRequest):
     """对历史比赛运行回测。
 
+    该接口会对每场已完赛比赛各发起一次 LLM 预测,成本高 —— 因此需要
+    `X-API-Key` 鉴权(见审查报告 P2-7)。
+
     对每场已完赛比赛:
       1. 用比赛之前的数据构建上下文 (防未来信息泄漏)
       2. 调 LLM 预测

src/api/routes/eval.py

@@ -5,6 +5,7 @@ import logging
 
 from fastapi import APIRouter, Depends, HTTPException
 
+from src.api.deps import require_admin_key
 from src.api.schemas import EvalSummaryOut, SettleRequest
 from src.db.base import AsyncSession, get_db, get_db_read
 from src.llm.eval import get_eval_summary, settle_prediction
@@ -14,7 +15,7 @@ logger = logging.getLogger(__name__)
 router = APIRouter(prefix="/api/v1", tags=["eval"])
 
 
-@router.post("/eval/settle")
+@router.post("/eval/settle", dependencies=[Depends(require_admin_key)])
 async def settle(req: SettleRequest, db: AsyncSession = Depends(get_db)):
     """回填实际结果。"""
     try:

src/api/routes/ingest.py

@@ -3,8 +3,9 @@ from __future__ import annotations
 
 import logging
 
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter, Depends, HTTPException
 
+from src.api.deps import require_admin_key
 from src.api.schemas import IngestBzzoiroRequest, IngestResponse, IngestUnderstatRequest, IngestInjuriesRequest, IngestSimpleResponse
 from src.data.sources import get_source
 from src.data.injuries import ingest_injuries
@@ -15,7 +16,7 @@ logger = logging.getLogger(__name__)
 router = APIRouter(prefix="/api/v1", tags=["ingest"])
 
 
-@router.post("/ingest/bzzoiro", response_model=IngestResponse)
+@router.post("/ingest/bzzoiro", response_model=IngestResponse, dependencies=[Depends(require_admin_key)])
 async def ingest_bzzoiro_route(req: IngestBzzoiroRequest):
     """触发 bzzoiro 采集。"""
     source = get_source("bzzoiro")
@@ -34,7 +35,7 @@ async def ingest_bzzoiro_route(req: IngestBzzoiroRequest):
         raise HTTPException(500, "数据采集失败,请查看服务器日志")
 
 
-@router.post("/ingest/understat", response_model=IngestSimpleResponse)
+@router.post("/ingest/understat", response_model=IngestSimpleResponse, dependencies=[Depends(require_admin_key)])
 async def ingest_understat_route(req: IngestUnderstatRequest):
     """触发 understat xG 回填。"""
     source = get_source("understat")
@@ -47,7 +48,7 @@ async def ingest_understat_route(req: IngestUnderstatRequest):
         raise HTTPException(500, "xG 回填失败,请查看服务器日志")
 
 
-@router.post("/ingest/injuries", response_model=IngestSimpleResponse)
+@router.post("/ingest/injuries", response_model=IngestSimpleResponse, dependencies=[Depends(require_admin_key)])
 async def ingest_injuries_route(req: IngestInjuriesRequest):
     """触发伤停采集。"""
     try:

src/core/config.py

@@ -32,6 +32,23 @@ class Settings(BaseSettings):
 
     # --- CORS ---
     CORS_ORIGINS: str = "http://localhost:5173,http://localhost:3000"
+    CORS_METHODS: str = "GET,POST,PUT,DELETE,OPTIONS"
+    CORS_HEADERS: str = "Authorization,Content-Type,X-API-Key,Accept"
+
+    # --- HTTP ---
+    HTTP_DEFAULT_TIMEOUT: int = 30
+
+    # --- database pool ---
+    DB_POOL_SIZE: int = 5
+    DB_MAX_OVERFLOW: int = 10
+    DB_POOL_TIMEOUT: int = 30
+    DB_POOL_RECYCLE: int = 1800
+
+    # --- 管理接口鉴权 ---
+    # 采集 / 回测等高成本或写入型接口需要此 Key(请求头 X-API-Key)。
+    # 留空表示「未启用鉴权」(本地开发默认),生产环境必须设置。
+    # 见审查报告 P2-7:ingest/backtest 无鉴权可被任意调用并烧掉 LLM 额度。
+    ADMIN_API_KEY: str = ""
 
 
 settings = Settings()

src/core/http_client.py

@@ -2,24 +2,27 @@
 
 使用方:
   - src/llm/provider.py: LLM 调用
+  - src/data/bzzoiro.py: bzzoiro 比赛数据
   - src/data/understat.py: xG 抓取
   - src/data/injuries.py: 伤停抓取
 
 生命周期由 FastAPI lifespan 管理(关闭时 aclose)。
+调用方可通过 `timeout` 参数覆盖 per-request 超时。
 """
 from __future__ import annotations
 
 import httpx
 
+from src.core.config import settings
+
 _shared_client: httpx.AsyncClient | None = None
-_default_timeout = 30
 
 
 def get_client() -> httpx.AsyncClient:
     """获取共享客户端(懒初始化)。"""
     global _shared_client
     if _shared_client is None or _shared_client.is_closed:
-        _shared_client = httpx.AsyncClient(timeout=_default_timeout)
+        _shared_client = httpx.AsyncClient(timeout=settings.HTTP_DEFAULT_TIMEOUT)
     return _shared_client
 
 

src/core/retry.py

@@ -1,82 +0,0 @@
-"""重试工具:带指数退避的瞬态错误重试。"""
-from __future__ import annotations
-
-import asyncio
-import functools
-import logging
-import random
-import time
-from typing import Callable, Iterable, TypeVar
-
-logger = logging.getLogger(__name__)
-
-T = TypeVar("T")
-
-
-def with_retry(
-    *,
-    max_retries: int = 3,
-    base_delay: float = 1.0,
-    max_delay: float = 30.0,
-    retryable_exceptions: Iterable[type[BaseException]] = (Exception,),
-    on_retry: Callable[[Exception, int], None] | None = None,
-) -> Callable:
-    """重试装饰器(同步/异步通用,指数退避 + 抖动)。
-
-    Args:
-        max_retries: 最大重试次数
-        base_delay: 基础延迟(秒)
-        max_delay: 最大延迟(秒)
-        retryable_exceptions: 触发重试的异常类型
-        on_retry: 重试回调(exception, attempt)
-    """
-    retryable = tuple(retryable_exceptions)
-
-    def decorator(func: Callable) -> Callable:
-        @functools.wraps(func)
-        async def async_wrapper(*args, **kwargs):
-            last_exc: Exception | None = None
-            for attempt in range(max_retries + 1):
-                try:
-                    return await func(*args, **kwargs)
-                except retryable as e:
-                    last_exc = e
-                    if attempt == max_retries:
-                        break
-                    delay = min(base_delay * (2 ** attempt), max_delay)
-                    delay += random.uniform(0, delay * 0.1)  # 抖动
-                    logger.warning(
-                        "%s failed (attempt %d/%d), retry in %.1fs: %s",
-                        func.__name__, attempt + 1, max_retries, delay, e,
-                    )
-                    if on_retry:
-                        on_retry(e, attempt + 1)
-                    await asyncio.sleep(delay)
-            raise last_exc  # type: ignore[misc]
-
-        @functools.wraps(func)
-        def sync_wrapper(*args, **kwargs):
-            last_exc: Exception | None = None
-            for attempt in range(max_retries + 1):
-                try:
-                    return func(*args, **kwargs)
-                except retryable as e:
-                    last_exc = e
-                    if attempt == max_retries:
-                        break
-                    delay = min(base_delay * (2 ** attempt), max_delay)
-                    delay += random.uniform(0, delay * 0.1)
-                    logger.warning(
-                        "%s failed (attempt %d/%d), retry in %.1fs: %s",
-                        func.__name__, attempt + 1, max_retries, delay, e,
-                    )
-                    if on_retry:
-                        on_retry(e, attempt + 1)
-                    time.sleep(delay)
-            raise last_exc  # type: ignore[misc]
-
-        if asyncio.iscoroutinefunction(func):
-            return async_wrapper
-        return sync_wrapper
-
-    return decorator

src/data/bzzoiro.py

@@ -2,6 +2,12 @@
 
 迁移自旧项目 app/data/sources/bzzoiro/,改成 async + 简化入库。
 使用 Repository 模式进行数据访问,不直接控制事务。
+
+改进(P1):
+  - Bronze 层集成:采集后先存 RawEvent,再规范化
+  - 死信表集成:采集/规范化失败写 IngestFailure
+  - 令牌桶限流:替换固定 REQUEST_INTERVAL sleep
+  - 数据血缘:记录 ETL 全过程
 """
 from __future__ import annotations
 
@@ -9,61 +15,98 @@ import asyncio
 import json as _json
 import logging
 import random
-import time as _time
-import urllib.error
-import urllib.parse
-import urllib.request
+import uuid
 from collections.abc import Iterable
 from datetime import datetime, timezone
 
 from sqlalchemy import select
 
 from src.core.config import settings
+from src.core.http_client import get_client
 from src.data.config import BZZOIRO_LEAGUE_IDS, LEAGUE_COUNTRIES, LEAGUE_NAMES, REQUEST_INTERVAL
 from src.data.normalize import normalize_bzzoiro
+from src.data.rate_limiter import RateLimitedClient, TokenBucket
 from src.data.sources import register
-from src.db.models import League, Match, MatchStats, Team
+from src.db.models import DataLineage, DataQualityCheck, IngestFailure, League, Match, MatchStats, RawEvent, Team
 
 logger = logging.getLogger(__name__)
 
+# 令牌桶限流器:替代固定 sleep,rate 根据 REQUEST_INTERVAL 计算
+_bzzoiro_limiter = TokenBucket(rate=1.0 / REQUEST_INTERVAL, capacity=3)
+# P2-3: RateLimitedClient 包装,在 HTTP 调用层面透明限流
+_bzzoiro_rl_client: RateLimitedClient | None = None
 
-def _fetch_json_sync(path: str, params: dict | None = None, max_retries: int = 3) -> dict | list:
-    """同步 HTTP(bzzoiro 客户端保持同步,在 async 函数里 run_in_executor)。"""
+
+def _get_bzzoiro_rl_client() -> RateLimitedClient:
+    """惰性初始化 RateLimitedClient(避免模块导入时创建 client)。"""
+    global _bzzoiro_rl_client
+    if _bzzoiro_rl_client is None:
+        client = get_client()
+        _bzzoiro_rl_client = RateLimitedClient(client, _bzzoiro_limiter)
+    return _bzzoiro_rl_client
+
+
+def _to_date(value):
+    """把 datetime / date / str 统一成 `date`。"""
+    if value is None:
+        return None
+    if hasattr(value, "date") and callable(value.date):
+        return value.date()
+    return value
+
+
+def _match_key(home_team_id: int, away_team_id: int, match_date) -> tuple[int, int, str]:
+    """比赛去重键:(主队, 客队, 天级日期 ISO 字符串)。
+
+    统一在这里构造,避免"预加载时用 str(date)、写入时用 isoformat()"这类
+    隐式格式依赖 —— 两者当前恰好相等,但一旦有人改动其一就会静默失配,
+    导致所有比赛被判为不存在而重复插入。
+    """
+    d = _to_date(match_date)
+    return (home_team_id, away_team_id, d.isoformat() if d is not None else "")
+
+
+async def _fetch_json_async(path: str, params: dict | None = None, max_retries: int = 3) -> dict | list:
+    """异步 HTTP(bzzoiro 使用 httpx,不再阻塞事件循环线程池)。"""
     base = settings.BZZOIRO_BASE.rstrip("/")
     url = f"{base}/{path.lstrip('/')}"
-    if params:
-        url += "?" + urllib.parse.urlencode(params)
     key = settings.BZZOIRO_KEY
     if not key:
         raise RuntimeError("BZZOIRO_KEY 未设置")
 
+    headers = {
+        "Authorization": f"Token {key}",
+        "Accept": "application/json",
+    }
+
     last_exc: Exception | None = None
     for attempt in range(max_retries):
         try:
-            req = urllib.request.Request(url)
-            req.add_header("Authorization", f"Token {key}")
-            req.add_header("Accept", "application/json")
-            with urllib.request.urlopen(req, timeout=30) as resp:
-                return _json.loads(resp.read().decode("utf-8"))
-        except urllib.error.HTTPError as e:
+            # P2-3: 使用 RateLimitedClient 包装 HTTP 调用,透明限流
+            rl_client = _get_bzzoiro_rl_client()
+            resp = await rl_client.get(url, headers=headers, params=params, timeout=30)
+            resp.raise_for_status()
+            return resp.json()
+        except Exception as e:
             last_exc = e
-            if e.code == 429:
-                # 指数退避: 429 通常意味着限速
+            status = getattr(getattr(e, "response", None), "status_code", None)
+            if status == 429:
                 delay = min(2 ** attempt, 16) + random.uniform(0, 1)
                 logger.warning("bzzoiro 429, retry %d in %.1fs", attempt + 1, delay)
-                _time.sleep(delay)
+                await asyncio.sleep(delay)
                 continue
-            if 500 <= e.code < 600:
+            if 500 <= (status or 0) < 600:
                 delay = min(2 ** attempt, 16) + random.uniform(0, 1)
-                logger.warning("bzzoiro %d, retry %d in %.1fs", e.code, attempt + 1, delay)
-                _time.sleep(delay)
+                logger.warning("bzzoiro %d, retry %d in %.1fs", status, attempt + 1, delay)
+                await asyncio.sleep(delay)
                 continue
-            raise  # 4xx 直接抛
-        except (urllib.error.URLError, TimeoutError, ConnectionError) as e:
-            last_exc = e
-            delay = min(2 ** attempt, 16) + random.uniform(0, 1)
-            logger.warning("bzzoiro network error, retry %d in %.1fs: %s", attempt + 1, delay, e)
-            _time.sleep(delay)
+            # 网络错误(连接失败/超时)也退避重试
+            if isinstance(e, (TimeoutError, ConnectionError, OSError)):
+                delay = min(2 ** attempt, 16) + random.uniform(0, 1)
+                logger.warning("bzzoiro network error, retry %d in %.1fs: %s", attempt + 1, delay, e)
+                await asyncio.sleep(delay)
+                continue
+            raise
     raise RuntimeError(f"bzzoiro request failed after {max_retries} attempts: {last_exc}")
 
 
@@ -75,15 +118,13 @@ async def fetch_bzzoiro_events(
     date_to: str | None = None,
     limit: int = 200,
 ) -> list[dict]:
-    """抓取 bzzoiro 原始事件(异步包装)。"""
+    """抓取 bzzoiro 原始事件(纯异步,无需 run_in_executor)。"""
     league_id = BZZOIRO_LEAGUE_IDS.get(league_code)
     if league_id is None:
         raise ValueError(f"未知联赛代码: {league_code}")
 
-    loop = asyncio.get_event_loop()
     rows: list[dict] = []
     offset = 0
-    payload: dict | list = {}
     while True:
         params: dict = {
             "league_id": league_id,
@@ -95,8 +136,7 @@ async def fetch_bzzoiro_events(
             params["date_from"] = str(date_from)[:10]
         if date_to:
             params["date_to"] = str(date_to)[:10]
-        # 显式位置参数,避免 lambda 闭包捕获循环变量
-        payload = await loop.run_in_executor(None, _fetch_json_sync, "/events/", params)
+        payload = await _fetch_json_async("/events/", params)
         batch = payload.get("results") or []
         if not batch:
             break
@@ -107,10 +147,141 @@ async def fetch_bzzoiro_events(
             break
         if len(batch) < limit:
             break
-        await asyncio.sleep(REQUEST_INTERVAL)
+        # 令牌桶已在 _fetch_json_async 内部处理,这里不再需要固定 sleep
     return rows
 
 
+async def _write_bronze_events(db, raw_events: list[dict], batch_id: str) -> None:
+    """将原始事件写入 Bronze 层(RawEvent 表)。
+
+    P2-1: 写入血缘记录,追踪 Bronze 层摄取过程。
+    """
+    for raw in raw_events:
+        try:
+            source_record_id = str(raw.get("id", ""))
+            if not source_record_id:
+                # P1-fix: 不再静默跳过,记录到死信表
+                logger.warning("bzzoiro raw event missing id: %s", str(raw)[:200])
+                await _write_ingest_failure(
+                    db, "bzzoiro", "match", "",
+                    "normalize_error", "raw event missing id", raw,
+                )
+                continue
+            bronze = RawEvent(
+                source_system="bzzoiro",
+                source_record_id=source_record_id,
+                raw_payload=raw,
+                ingest_batch_id=batch_id,
+            )
+            db.add(bronze)
+        except Exception as e:
+            logger.warning("bzzoiro raw_event write failed: %s", e)
+    try:
+        await db.flush()
+    except Exception as e:
+        logger.warning("bzzoiro raw_events flush failed: %s", e)
+
+    # P2-1: 写入 Bronze 层血缘记录
+    for raw in raw_events:
+        source_record_id = str(raw.get("id", ""))
+        if not source_record_id:
+            continue
+        await _write_data_lineage(
+            db,
+            source_system="bzzoiro",
+            source_record_id=source_record_id,
+            target_table="raw_events",
+            transform_name="bronze_ingest",
+            batch_id=batch_id,
+        )
+
+
+async def _write_ingest_failure(
+    db,
+    source_system: str,
+    entity_type: str,
+    source_record_id: str,
+    error_type: str,
+    error_detail: str,
+    raw_payload: dict | None,
+) -> None:
+    """写入采集失败到死信表(IngestFailure)。"""
+    try:
+        from datetime import timedelta
+        failure = IngestFailure(
+            source_system=source_system,
+            entity_type=entity_type,
+            source_record_id=source_record_id,
+            error_type=error_type,
+            error_detail=error_detail[:2000] if error_detail else None,
+            raw_payload=raw_payload,
+            status="pending",
+            next_retry_at=datetime.now(timezone.utc) + timedelta(minutes=5),
+        )
+        db.add(failure)
+        await db.flush()
+    except Exception:
+        logger.warning("bzzoiro ingest_failure write failed", exc_info=True)
+
+
+async def _write_data_lineage(
+    db,
+    *,
+    source_system: str,
+    source_record_id: str,
+    target_table: str,
+    target_id: int | None = None,
+    transform_name: str,
+    transform_detail: str | None = None,
+    batch_id: str | None = None,
+) -> None:
+    """写入数据血缘记录(DataLineage 表)。"""
+    try:
+        lineage = DataLineage(
+            source_system=source_system,
+            source_record_id=source_record_id,
+            target_table=target_table,
+            target_id=target_id,
+            transform_name=transform_name,
+            transform_detail=transform_detail,
+            batch_id=batch_id,
+        )
+        db.add(lineage)
+        await db.flush()
+    except Exception:
+        logger.warning("bzzoiro data_lineage write failed", exc_info=True)
+
+
+async def _write_data_quality_check(
+    db,
+    *,
+    check_name: str,
+    entity_type: str,
+    entity_id: str | None = None,
+    expected_value: str | None = None,
+    actual_value: str | None = None,
+    passed: bool,
+    severity: str = "warning",
+    detail: str | None = None,
+) -> None:
+    """写入数据质量检查结果(DataQualityCheck 表)。"""
+    try:
+        check = DataQualityCheck(
+            check_name=check_name,
+            entity_type=entity_type,
+            entity_id=entity_id,
+            expected_value=expected_value,
+            actual_value=actual_value,
+            passed=passed,
+            severity=severity,
+            detail=detail,
+        )
+        db.add(check)
+        await db.flush()
+    except Exception:
+        logger.warning("bzzoiro data_quality_check write failed", exc_info=True)
+
+
 @register
 class BzzoiroSource:
     """bzzoiro 数据源(实现 DataSource 协议)。"""
@@ -131,6 +302,7 @@ class BzzoiroSource:
         注意: 本方法不控制事务(commit/rollback),由调用方通过 UnitOfWork 控制。
         """
         result: dict = {"leagues": {}, "total_inserted": 0, "total_updated": 0, "errors": []}
+        batch_id = uuid.uuid4().hex[:16]
 
         for code in leagues:
             league_r: dict = {"inserted": 0, "updated": 0, "errors": []}
@@ -139,9 +311,17 @@ class BzzoiroSource:
             except Exception as e:
                 logger.exception("bzzoiro fetch failed for %s", code)
                 league_r["errors"].append(f"fetch failed: {e}")
+                # 写死信表
+                await _write_ingest_failure(
+                    db, "bzzoiro", "match", f"fetch_{code}_{batch_id}",
+                    "fetch_error", str(e), {"league": code, "date_from": date_from, "date_to": date_to},
+                )
                 result["leagues"][code] = league_r
                 continue
 
+            # === Bronze 层:先存 RawEvent ===
+            await _write_bronze_events(db, raw_events, batch_id)
+
             # 获取或创建联赛
             stmt = select(League).where(League.code == code)
             league = (await db.execute(stmt)).scalar_one_or_none()
@@ -153,7 +333,9 @@ class BzzoiroSource:
             # === 批量优化: 预加载球队和已有比赛到内存 ===
             team_name_to_id: dict[str, int] = {}
             existing_matches: dict[tuple[int, int, str], Match] = {}  # 完整对象,避免重复查询
-            normalized_matches: list = []  # 缓存规范化结果,避免重复调用
+            # (NormalizedMatch, 原始 event) 成对保存:后续写 source_record_id 时
+            # 必须用配对的那条 event,不能依赖外层循环变量残留值。
+            normalized_matches: list[tuple] = []
 
             if raw_events:
                 # 一次遍历: 收集球队名 + 规范化
@@ -164,10 +346,15 @@ class BzzoiroSource:
                         try:
                             nm.validate()
                         except Exception as e:
-                            logger.debug("normalize skip: %s", e)
-                            league_r["errors"].append(f"normalize: {e}")
+                            # P1-3: 统一使用 warning,不追加到 errors(仅运行时错误入 errors)
+                            logger.warning("normalize skip: %s", e)
+                            # 写死信表:规范化失败
+                            await _write_ingest_failure(
+                                db, "bzzoiro", "match", str(raw.get("id", "")),
+                                "normalize_error", str(e), raw,
+                            )
                             continue
-                        normalized_matches.append(nm)
+                        normalized_matches.append((nm, raw))
                         all_team_names.add(nm.home_team)
                         all_team_names.add(nm.away_team)
 
@@ -176,13 +363,27 @@ class BzzoiroSource:
                     teams = (await db.execute(stmt)).scalars().all()
                     team_name_to_id = {t.name: t.id for t in teams}
 
-                # 预加载已有比赛(完整对象)
-                stmt = select(Match).where(Match.league_id == league.id)
-                for m in (await db.execute(stmt)).scalars():
-                    key = (m.home_team_id, m.away_team_id, str(m.match_date_date))
-                    existing_matches[key] = m
+                # P1-2: 按需加载,只加载 raw_events 涉及日期范围的比赛(加 30 天缓冲)
+                # 避免加载联赛全部历史比赛到内存(多赛季采集时内存溢出)
+                if normalized_matches:
+                    from datetime import timedelta
+                    dates = [nm.date for nm in normalized_matches if nm.date is not None]
+                    if dates:
+                        min_dt = min(dates) - timedelta(days=30)
+                        max_dt = max(dates) + timedelta(days=30)
+                        stmt = (
+                            select(Match)
+                            .where(Match.league_id == league.id)
+                            .where(Match.match_date >= min_dt)
+                            .where(Match.match_date <= max_dt)
+                        )
+                        existing_matches = {
+                            _match_key(m.home_team_id, m.away_team_id, m.match_date_date): m
+                            for m in (await db.execute(stmt)).scalars()
+                        }
+                # else: existing_matches 保持空 dict(全量新比赛)
 
-            for nm in normalized_matches:
+            for nm, raw in normalized_matches:
                 # 球队: 内存查找 + 按需创建
                 home_team_id = team_name_to_id.get(nm.home_team)
                 if home_team_id is None:
@@ -201,8 +402,7 @@ class BzzoiroSource:
                     team_name_to_id[nm.away_team] = away_team_id
 
                 # 查找已有比赛: 内存查找
-                date_key = nm.date.date().isoformat() if hasattr(nm.date, "date") else str(nm.date)
-                match_key = (home_team_id, away_team_id, date_key)
+                match_key = _match_key(home_team_id, away_team_id, nm.date)
                 existing_match = existing_matches.get(match_key)
 
                 if existing_match is None:
@@ -212,7 +412,7 @@ class BzzoiroSource:
                         home_team_id=home_team_id,
                         away_team_id=away_team_id,
                         match_date=nm.date,
-                        match_date_date=nm.date.date() if hasattr(nm.date, "date") else nm.date,
+                        match_date_date=_to_date(nm.date),
                         match_status=nm.match_status,
                         home_goals=nm.home_goals,
                         away_goals=nm.away_goals,
@@ -223,6 +423,16 @@ class BzzoiroSource:
                     db.add(m)
                     await db.flush()
                     existing_matches[match_key] = m  # 防止同批重复
+                    # P2-1: Silver 层血缘 — Match 创建
+                    await _write_data_lineage(
+                        db,
+                        source_system="bzzoiro",
+                        source_record_id=str(raw.get("id", "")),
+                        target_table="matches",
+                        target_id=m.id,
+                        transform_name="silver_upsert",
+                        batch_id=batch_id,
+                    )
                     if nm.home_xg is not None or nm.away_xg is not None:
                         now = datetime.now(timezone.utc)
                         stats = MatchStats(
@@ -241,11 +451,22 @@ class BzzoiroSource:
                             home_red_cards=nm.home_red_cards,
                             away_red_cards=nm.away_red_cards,
                             source="bzzoiro",
-                            source_event_id=str(raw.get("id", "")),
+                            source_record_id=str(raw.get("id", "")),
                             retrieved_at=now,
                             available_at=now,
                         )
                         db.add(stats)
+                        await db.flush()
+                        # P2-1: Silver 层血缘 — MatchStats 创建
+                        await _write_data_lineage(
+                            db,
+                            source_system="bzzoiro",
+                            source_record_id=str(raw.get("id", "")),
+                            target_table="match_stats",
+                            target_id=m.id,
+                            transform_name="silver_upsert",
+                            batch_id=batch_id,
+                        )
                     league_r["inserted"] += 1
                 else:
                     # 已有比赛: 直接从内存获取对象更新(无需再查询)
@@ -263,9 +484,39 @@ class BzzoiroSource:
                         existing_match.match_stage = nm.match_stage
                         changed = True
                     if existing_match.stats is None and (nm.home_xg is not None or nm.away_xg is not None):
-                        existing_match.stats = MatchStats(match_id=existing_match.id)
+                        now = datetime.now(timezone.utc)
+                        existing_match.stats = MatchStats(
+                            match_id=existing_match.id,
+                            home_xg=nm.home_xg,
+                            away_xg=nm.away_xg,
+                            home_shots=nm.home_shots,
+                            away_shots=nm.away_shots,
+                            home_shots_on_target=nm.home_shots_on_target,
+                            away_shots_on_target=nm.away_shots_on_target,
+                            home_corners=nm.home_corners,
+                            away_corners=nm.away_corners,
+                            home_possession=nm.home_possession,
+                            home_yellow_cards=nm.home_yellow_cards,
+                            away_yellow_cards=nm.away_yellow_cards,
+                            home_red_cards=nm.home_red_cards,
+                            away_red_cards=nm.away_red_cards,
+                            source="bzzoiro",
+                            source_record_id=str(raw.get("id", "")),
+                            retrieved_at=now,
+                            available_at=now,
+                        )
                         db.add(existing_match.stats)
                         await db.flush()
+                        # P2-1: Silver 层血缘 — MatchStats 创建(更新路径)
+                        await _write_data_lineage(
+                            db,
+                            source_system="bzzoiro",
+                            source_record_id=str(raw.get("id", "")),
+                            target_table="match_stats",
+                            target_id=existing_match.id,
+                            transform_name="silver_upsert",
+                            batch_id=batch_id,
+                        )
                     if existing_match.stats is not None:
                         for fld in ("home_xg", "away_xg", "home_shots", "away_shots",
                                     "home_shots_on_target", "away_shots_on_target",
@@ -284,4 +535,19 @@ class BzzoiroSource:
             result["leagues"][code] = league_r
             result["total_inserted"] += league_r["inserted"]
             result["total_updated"] += league_r["updated"]
+
+            # P2-2: 数据质量检查 — 行计数合理性(某联赛比赛数不应为 0)
+            total_records = league_r["inserted"] + league_r["updated"]
+            await _write_data_quality_check(
+                db,
+                check_name="bzzoiro_row_count",
+                entity_type="match",
+                entity_id=code,
+                expected_value=">=1",
+                actual_value=str(total_records),
+                passed=total_records > 0,
+                severity="critical" if total_records == 0 else "info",
+                detail=f"league={code} inserted={league_r['inserted']} updated={league_r['updated']}",
+            )
+
         return result

src/data/injuries.py

@@ -1,15 +1,25 @@
 """伤停数据采集器(api-football / api-sports.io)。
 
 采集伤停数据并入库(injuries 表),供 injuries agent 使用。
+
+改进(P0):
+  - 原子写缓存(tempfile + os.replace)
+  - TTL 分级:未来比赛 1h,当天 5min,历史 7day
+  - 文件锁防止并发写缓存冲突
+  - Bronze 层集成(RawEvent)
+  - 死信表集成(IngestFailure)
 """
 from __future__ import annotations
 
 import asyncio
 import json
 import logging
+import os
 import random
+import tempfile
 import time
-from datetime import datetime, timezone
+import uuid
+from datetime import datetime, timedelta, timezone
 from pathlib import Path
 from typing import Any
 
@@ -17,14 +27,147 @@ import httpx
 
 from src.core.config import settings
 from src.core.http_client import get_client
+from src.db.models import DataQualityCheck
 
 logger = logging.getLogger(__name__)
 
 API_BASE = "https://v3.football.api-sports.io"
 DEFAULT_HOST = "v3.football.api-sports.io"
 
-# 缓存目录
-_CACHE_DIR = Path(__file__).resolve().parent.parent.parent / "data" / "injuries_cache"
+# P2-3: 缓存目录改用系统临时目录,避免源码树内写入
+_CACHE_DIR = Path(tempfile.gettempdir()) / "profeto_injuries"
+
+# 模块级锁:防止并发写同一缓存文件
+_cache_lock = asyncio.Lock()
+
+
+def _sanitize_cache_component(value: str | int | None) -> str:
+    """消毒缓存键组件,防止路径遍历。
+
+    只允许 [A-Za-z0-9_.-],其余字符替换为 '_'。
+    """
+    import re
+    s = str(value) if value is not None else "none"
+    return re.sub(r'[^A-Za-z0-9_.]', '_', s)[:60]
+
+
+def _build_cache_key(prefix: str, *components: str | int | None) -> str:
+    """构造安全的缓存文件名。"""
+    parts = [_sanitize_cache_component(c) for c in components]
+    return f"{prefix}_{'_'.join(parts)}.json"
+
+
+def _compute_ttl_hours(date_str: str | None, fixture_id: int | None, league_id: int | None) -> float:
+    """根据查询参数计算缓存 TTL(小时)。
+
+    TTL 分级策略:
+      - 未来比赛(date > now): 1 小时(赛前伤停变化频繁)
+      - 当天比赛(date == today): 5 分钟(赛中实时更新)
+      - 历史比赛(date < now): 7 天(历史数据不变)
+      - 无日期参数: 1 小时(保守策略)
+    """
+    if date_str is None:
+        # 无日期参数(按 fixture_id 或 league_id 查询),保守 TTL
+        return 1.0
+    try:
+        query_date = datetime.strptime(date_str, "%Y-%m-%d").replace(tzinfo=timezone.utc)
+        now = datetime.now(timezone.utc)
+        today = now.date()
+        if query_date.date() > today:
+            return 1.0  # 未来
+        elif query_date.date() == today:
+            return 5.0 / 60  # 当天:5 分钟
+        else:
+            return 168.0  # 历史:7 天
+    except (ValueError, TypeError):
+        return 1.0  # 解析失败,保守 TTL
+
+
+def _write_cache_atomic(cache_file: Path, data: Any) -> None:
+    """原子写缓存文件。
+
+    使用 tempfile + os.replace 实现原子写,防止读到写了一半的文件。
+    配合模块级 asyncio.Lock,杜绝并发写冲突。
+    """
+    cache_file.parent.mkdir(parents=True, exist_ok=True)
+    # 在同一目录创建临时文件(保证 os.replace 是原子操作)
+    fd, tmp_path = tempfile.mkstemp(
+        dir=str(cache_file.parent),
+        prefix=f".{cache_file.name}.tmp_",
+    )
+    try:
+        with os.fdopen(fd, "w", encoding="utf-8") as f:
+            json.dump(data, f, ensure_ascii=False, default=str)
+        os.replace(tmp_path, cache_file)
+    except BaseException:
+        # 失败时清理临时文件
+        try:
+            os.unlink(tmp_path)
+        except OSError:
+            pass
+        raise
+
+
+def _cleanup_expired_cache() -> int:
+    """P2-4: 扫描缓存目录,删除过期的 .json 文件。
+
+    TTL 策略(保守取最大值,避免误删有效缓存):
+      - 最短 TTL 为 5 分钟(当天比赛),但清理阈值用 1 小时
+      - 超过 1 小时的缓存文件视为过期
+
+    Returns:
+        删除的文件数。
+    """
+    import glob as _glob
+
+    cache_dir = _CACHE_DIR
+    if not cache_dir.exists():
+        return 0
+
+    removed = 0
+    max_ttl_seconds = 3600  # 1 小时(保守阈值,最短实际 TTL 5 分钟)
+    now_ts = time.time()
+    for cache_file in _glob.glob(str(cache_dir / "*.json")):
+        try:
+            file_age = now_ts - os.path.getmtime(cache_file)
+            if file_age > max_ttl_seconds:
+                os.unlink(cache_file)
+                removed += 1
+        except OSError:
+            continue
+    if removed > 0:
+        logger.info("injuries cache cleanup: removed %d expired files", removed)
+    return removed
+
+
+async def _write_data_quality_check(
+    db,
+    *,
+    check_name: str,
+    entity_type: str,
+    entity_id: str | None = None,
+    expected_value: str | None = None,
+    actual_value: str | None = None,
+    passed: bool,
+    severity: str = "warning",
+    detail: str | None = None,
+) -> None:
+    """写入数据质量检查结果(DataQualityCheck 表)。"""
+    try:
+        check = DataQualityCheck(
+            check_name=check_name,
+            entity_type=entity_type,
+            entity_id=entity_id,
+            expected_value=expected_value,
+            actual_value=actual_value,
+            passed=passed,
+            severity=severity,
+            detail=detail,
+        )
+        db.add(check)
+        await db.flush()
+    except Exception:
+        logger.warning("injuries data_quality_check write failed", exc_info=True)
 
 
 async def fetch_injuries(*, date: str | None = None, fixture_id: int | None = None, league_id: int | None = None) -> list[dict]:
@@ -42,20 +185,25 @@ async def fetch_injuries(*, date: str | None = None, fixture_id: int | None = No
     if not api_key:
         raise RuntimeError("API_FOOTBALL_KEY 未设置")
 
+    # P2-4: 清理过期缓存文件
+    _cleanup_expired_cache()
+
     cache_dir = _CACHE_DIR
     cache_dir.mkdir(parents=True, exist_ok=True)
 
-    # 缓存命中 (7 天内有效)
-    cache_key = f"injuries_{date}_{fixture_id}_{league_id}.json"
+    # TTL 分级缓存
+    ttl_hours = _compute_ttl_hours(date, fixture_id, league_id)
+    # P0-fix: 消毒缓存键,防止路径遍历(date 参数可能含 '../' 等)
+    cache_key = _build_cache_key("injuries", date, fixture_id, league_id)
     cache_file = cache_dir / cache_key
     if cache_file.exists():
         age_hours = (time.time() - cache_file.stat().st_mtime) / 3600
-        if age_hours < 168:  # 7 天
-            logger.debug("injuries cache hit: %s (%.1fh old)", cache_key, age_hours)
+        if age_hours < ttl_hours:
+            logger.debug("injuries cache hit: %s (%.1fh old, ttl=%.2fh)", cache_key, age_hours, ttl_hours)
             with open(cache_file, encoding="utf-8") as f:
                 return json.load(f)
         else:
-            logger.debug("injuries cache expired: %s (%.1fh old)", cache_key, age_hours)
+            logger.debug("injuries cache expired: %s (%.1fh old, ttl=%.2fh)", cache_key, age_hours, ttl_hours)
 
     headers = {
         "x-apisports-key": api_key,
@@ -92,9 +240,9 @@ async def fetch_injuries(*, date: str | None = None, fixture_id: int | None = No
     data = resp.json()
     injuries = data.get("response", [])
 
-    # 写缓存
-    with open(cache_file, "w", encoding="utf-8") as f:
-        json.dump(injuries, ensure_ascii=False, default=str, fp=f)
+    # 原子写缓存 + 文件锁(防止并发写冲突)
+    async with _cache_lock:
+        _write_cache_atomic(cache_file, injuries)
 
     return injuries
 
@@ -103,27 +251,74 @@ async def ingest_injuries(db, *, date: str | None = None) -> dict:
     """采集伤停数据并入库(injuries 表)。
 
     注意: 本方法不控制事务(commit/rollback),由调用方通过 UnitOfWork 控制。
+
+    P1-4: 批量幂等检查,避免逐条查询的竞态条件(并发采集时 IntegrityError)。
+    P1 Bronze: 采集成功后先存 RawEvent,再规范化。
+    P1 死信: 采集/规范化失败时写 IngestFailure。
     """
     from sqlalchemy import select
+    from sqlalchemy.exc import IntegrityError
+    from sqlalchemy.orm import selectinload
 
     from src.data.team_names import normalize as normalize_name
-    from src.db.models import Injury, Team
+    from src.db.models import DataQualityCheck, Injury, IngestFailure, RawEvent, Team
 
     result = {"count": 0, "inserted": 0, "errors": []}
+    batch_id = uuid.uuid4().hex[:16]
 
     try:
         raw_injuries = await fetch_injuries(date=date)
     except Exception as e:
         logger.exception("injuries fetch failed")
         result["errors"].append(f"fetch failed: {e}")
+        # 写死信表(IngestFailure)
+        try:
+            failure = IngestFailure(
+                source_system="api-football",
+                entity_type="injury",
+                source_record_id=f"fetch_{batch_id}",
+                error_type="fetch_error",
+                error_detail=str(e)[:2000],
+                raw_payload={"date": date},
+                status="pending",
+                next_retry_at=datetime.now(timezone.utc) + timedelta(minutes=5),
+            )
+            db.add(failure)
+            await db.flush()
+        except Exception:
+            logger.warning("failed to write fetch error to ingest_failures", exc_info=True)
         return result
 
     result["count"] = len(raw_injuries)
 
+    # === Bronze 层:写 RawEvent ===
+    for raw in raw_injuries:
+        try:
+            player = raw.get("player", {}) or {}
+            fixture = raw.get("fixture", {}) or {}
+            source_record_id = f"inj_{player.get('id')}_{fixture.get('id')}"
+            bronze = RawEvent(
+                source_system="api-football",
+                source_record_id=source_record_id,
+                raw_payload=raw,
+                ingest_batch_id=batch_id,
+            )
+            db.add(bronze)
+        except Exception as e:
+            logger.warning("raw_event write failed for injury: %s", e)
+    try:
+        await db.flush()
+    except Exception as e:
+        logger.warning("raw_events flush failed: %s", e)
+
     # 预加载所有球队(用于按名匹配)
     teams = (await db.execute(select(Team))).scalars().all()
     team_by_name = {t.name: t.id for t in teams}
 
+    # P1-4: 收集所有待插入记录的键,批量查询已存在的记录
+    # 避免逐条查询 + 插入的竞态条件(两个并发请求同时通过检查 → IntegrityError)
+    pending_records: list[dict] = []
+    parse_failures: list[dict] = []  # 规范化失败的原始数据,用于写死信表
     for raw in raw_injuries:
         try:
             player = raw.get("player", {}) or {}
@@ -144,43 +339,145 @@ async def ingest_injuries(db, *, date: str | None = None) -> dict:
                 except (ValueError, AttributeError):
                     pass
 
+            # 强制 int 转换,API 可能返回字符串
             player_id = player.get("id")
+            try:
+                player_id = int(player_id) if player_id is not None else None
+            except (ValueError, TypeError):
+                player_id = None
             fixture_id = fixture.get("id")
+            try:
+                fixture_id = int(fixture_id) if fixture_id is not None else None
+            except (ValueError, TypeError):
+                fixture_id = None
 
-            # 幂等: 已存在则跳过
-            existing = (
-                await db.execute(
-                    select(Injury).where(
-                        Injury.player_id == player_id,
-                        Injury.fixture_id == fixture_id,
-                        Injury.injury_type == player.get("type"),
-                    )
-                )
-            ).scalar_one_or_none()
-
-            if existing is not None:
-                continue
-
-            injury = Injury(
-                player_id=player_id,
-                player_name=player_name,
-                team_id=team_id,
-                fixture_id=fixture_id,
-                league_id=(raw.get("league") or {}).get("id"),
-                injury_type=player.get("type"),
-                reason=player.get("reason"),
-                injury_date=injury_date,
-            )
-            db.add(injury)
-            result["inserted"] += 1
+            pending_records.append({
+                "player_id": player_id,
+                "player_name": player_name,
+                "team_id": team_id,
+                "fixture_id": fixture_id,
+                "league_id": (raw.get("league") or {}).get("id"),
+                "injury_type": player.get("type"),
+                "reason": player.get("reason"),
+                "injury_date": injury_date,
+            })
         except Exception as e:
             result["errors"].append(f"parse error: {e}")
+            parse_failures.append({"raw": raw, "error": str(e)})
+
+    # 写规范化失败到死信表
+    for fail in parse_failures:
+        try:
+            raw = fail["raw"]
+            player = raw.get("player", {}) or {}
+            failure = IngestFailure(
+                source_system="api-football",
+                entity_type="injury",
+                source_record_id=f"inj_{player.get('id', 'unknown')}",
+                error_type="normalize_error",
+                error_detail=fail["error"][:2000],
+                raw_payload=raw,
+                status="pending",
+                next_retry_at=datetime.now(timezone.utc) + timedelta(minutes=5),
+            )
+            db.add(failure)
+        except Exception:
+            logger.warning("failed to write parse error to ingest_failures", exc_info=True)
+    if parse_failures:
+        try:
+            await db.flush()
+        except Exception:
+            logger.warning("ingest_failures flush failed", exc_info=True)
+
+    # P1-4: 批量查询已存在的记录(1 次 DB 往返)
+    existing_keys: set[tuple] = set()
+    if pending_records:
+        # 构造查询条件:所有 (player_id, fixture_id, injury_type) 组合
+        # 使用 OR 条件批量查询
+        conditions = []
+        for rec in pending_records:
+            conditions.append(
+                (Injury.player_id == rec["player_id"])
+                & (Injury.fixture_id == rec["fixture_id"])
+                & (Injury.injury_type == rec["injury_type"])
+            )
+        if conditions:
+            from sqlalchemy import or_
+            stmt = select(Injury.player_id, Injury.fixture_id, Injury.injury_type).where(or_(*conditions))
+            rows = (await db.execute(stmt)).all()
+            existing_keys = {(r[0], r[1], r[2]) for r in rows}
+
+    # P1-4: 批量插入(跳过已存在的)
+    for rec in pending_records:
+        key = (rec["player_id"], rec["fixture_id"], rec["injury_type"])
+        if key in existing_keys:
+            continue
+
+        injury = Injury(**rec)
+        db.add(injury)
+        result["inserted"] += 1
+
+        # 每 50 条 flush 一次,减少内存压力,同时捕获 IntegrityError
+        if result["inserted"] % 50 == 0:
+            try:
+                await db.flush()
+            except IntegrityError:
+                # P1-4: 并发采集时可能仍有竞态,回退到逐条插入
+                await db.rollback()
+                logger.warning("injuries batch IntegrityError, falling back to per-record insert")
+                return await _ingest_injuries_fallback(db, pending_records, result)
+
+    # 最终 flush
+    try:
+        await db.flush()
+    except IntegrityError:
+        await db.rollback()
+        logger.warning("injuries final flush IntegrityError, falling back to per-record insert")
+        return await _ingest_injuries_fallback(db, pending_records, result)
+
+    # P2-2: 数据质量检查 — 行计数 / 新增比例
+    total = result["count"]
+    inserted = result["inserted"]
+    # 某日期伤停数不应为 0(除非历史日期)
+    await _write_data_quality_check(
+        db,
+        check_name="injuries_row_count",
+        entity_type="injury",
+        entity_id=date or "unknown",
+        expected_value=">=0",
+        actual_value=str(total),
+        passed=True,
+        severity="info",
+        detail=f"date={date} fetched={total} inserted={inserted}",
+    )
 
     # 注意: 不在此处 commit,由调用方 UnitOfWork 控制事务
     logger.info("injuries: fetched %d, inserted %d for %s", result["count"], result["inserted"], date)
     return result
 
 
+async def _ingest_injuries_fallback(db, pending_records: list[dict], result: dict) -> dict:
+    """P1-4: 逐条插入回退,捕获每条 IntegrityError 避免整批回滚。"""
+    from sqlalchemy.exc import IntegrityError
+    from src.db.models import Injury
+
+    inserted = 0
+    for rec in pending_records:
+        injury = Injury(**rec)
+        db.add(injury)
+        try:
+            await db.flush()
+            inserted += 1
+        except IntegrityError:
+            await db.rollback()
+            # 已存在或其他冲突,跳过
+            continue
+
+    result["inserted"] = inserted
+    logger.info("injuries fallback: inserted %d records", inserted)
+    return result
+
+
 async def get_injuries_for_match(db, team_id: int, match_date, as_of=None) -> list[Injury]:
     """查询某场比赛前某队的伤停名单(比赛日仍缺阵的)。
 
@@ -188,31 +485,29 @@ async def get_injuries_for_match(db, team_id: int, match_date, as_of=None) -> li
         db: 数据库 session
         team_id: 球队 ID
         match_date: 比赛日期
-        as_of: 截止时间(cutoff)。只返回 retrieved_at <= as_of 的记录。
-               用于回测时防止"未来采集的数据"泄漏到历史预测。
-               必须保持 timezone-aware datetime,不会截断为 date。
-    """
-    from sqlalchemy import or_, select
+        as_of: 数据截止时间(用于回测防泄漏)
 
+    Returns:
+        伤停记录列表
+    """
+    from sqlalchemy import select
     from src.db.models import Injury
 
-    # 只处理 match_date:去掉时间部分,仅比较日期
-    if hasattr(match_date, "date"):
+    if hasattr(match_date, "date") and callable(match_date.date):
         match_date = match_date.date()
 
     stmt = (
         select(Injury)
         .where(Injury.team_id == team_id)
         .where(Injury.injury_date <= match_date)
-        .where(or_(Injury.return_date.is_(None), Injury.return_date >= match_date))
+        .where(
+            (Injury.return_date.is_(None)) | (Injury.return_date >= match_date)
+        )
     )
-
-    # 回测防泄漏: 只使用 as_of 时间点之前已采集的数据
-    # 注意: as_of 保持 datetime,不截断为 date,避免错误排除同日合法数据
     if as_of is not None:
-        stmt = stmt.where(Injury.retrieved_at.is_not(None))
+        if hasattr(as_of, "date") and callable(as_of.date):
+            as_of = as_of.date()
         stmt = stmt.where(Injury.retrieved_at <= as_of)
 
-    stmt = stmt.order_by(Injury.injury_date.desc())
     result = await db.execute(stmt)
     return list(result.scalars().all())

src/data/normalize.py

@@ -1,9 +1,9 @@
 """数据规范化:任意数据源原始记录 → NormalizedMatch。
 
 迁移自旧项目 app/data/normalize.py,简化:
-  - 去掉 XGBackfill 双轨(不再需要独立回填)
-  - 去掉 PIT 时间契约(无训练集要防泄漏)
-  - 保留核心清洗契约(队名归一、日期解析、数值范围)
+    - 去掉 XGBackoff 双轨(不再需要独立回填)
+    - 去掉 PIT 时间契约(无训练集要防泄漏)
+    - 保留核心清洗契约(队名归一、日期解析、数值范围)
 """
 from __future__ import annotations
 
@@ -90,7 +90,10 @@ def derive_season_label(date: datetime) -> str:
 
 
 def _parse_date(value) -> datetime | None:
-    """日期解析 → UTC datetime(带 tzinfo)。"""
+    """日期解析 → UTC datetime(带 tzinfo)。
+
+    P2-2: 解析失败时记录 warning,避免静默丢数据而无感知。
+    """
     if value in (None, ""):
         return None
     if isinstance(value, (int, float)):
@@ -111,6 +114,8 @@ def _parse_date(value) -> datetime | None:
             return datetime.strptime(s[:19], fmt).replace(tzinfo=timezone.utc)
         except ValueError:
             continue
+    # P2-2 修复: 记录被丢弃的原始值,便于排查数据源格式变更
+    logger.warning("_parse_date failed, dropping record: %r", value)
     return None
 
 
@@ -204,6 +209,6 @@ def normalize_understat(raw: dict, league_type: str) -> NormalizedMatch | None:
         away_team=away,
         match_status="finished",
         season_label=derive_season_label(dt),
-        home_xg=_to_float(home_xg),
-        away_xg=_to_float(away_xg),
+        home_xg=home_xg,
+        away_xg=away_xg,
     )

src/data/rate_limiter.py

@@ -0,0 +1,98 @@
+"""令牌桶限流器。
+
+替代 bzzoiro 中固定的 REQUEST_INTERVAL sleep,提供更精细的速率控制。
+支持突发流量和平滑限流。
+"""
+from __future__ import annotations
+
+import asyncio
+import logging
+import time
+
+logger = logging.getLogger(__name__)
+
+
+class TokenBucket:
+    """异步令牌桶限流器。
+
+    用法:
+        limiter = TokenBucket(rate=5.0, capacity=10)
+        await limiter.acquire()  # 等待直到有可用令牌
+
+    Args:
+        rate: 每秒补充的令牌数
+        capacity: 桶容量(允许的最大突发量)
+    """
+
+    def __init__(self, rate: float, capacity: int | None = None) -> None:
+        if rate <= 0:
+            raise ValueError(f"rate must be positive, got {rate}")
+        self._rate = rate
+        self._capacity = capacity or max(1, int(rate * 2))
+        self._tokens: float = self._capacity
+        self._last_refill = time.monotonic()
+        self._lock = asyncio.Lock()
+
+    async def acquire(self, tokens: int = 1) -> None:
+        """获取指定数量的令牌,不足时等待。
+
+        Args:
+            tokens: 需要获取的令牌数
+        """
+        if tokens <= 0:
+            return
+        if tokens > self._capacity:
+            raise ValueError(f"requested {tokens} exceeds capacity {self._capacity}")
+
+        while True:
+            async with self._lock:
+                self._refill()
+                if self._tokens >= tokens:
+                    self._tokens -= tokens
+                    return
+
+            # 计算需要等待的时间
+            wait_time = (tokens - self._tokens) / self._rate
+            logger.debug("token bucket: waiting %.2fs for %d tokens", wait_time, tokens)
+            await asyncio.sleep(wait_time)
+
+    def _refill(self) -> None:
+        """补充令牌(基于经过的时间)。"""
+        now = time.monotonic()
+        elapsed = now - self._last_refill
+        if elapsed > 0:
+            self._tokens = min(self._capacity, self._tokens + elapsed * self._rate)
+            self._last_refill = now
+
+    @property
+    def tokens(self) -> float:
+        """当前可用令牌数(近似)。"""
+        self._refill()
+        return self._tokens
+
+
+class RateLimitedClient:
+    """HTTP 客户端限流包装器。
+
+    在 httpx 客户端之上添加令牌桶限流,透明地控制请求速率。
+
+    用法:
+        client = get_client()
+        limiter = TokenBucket(rate=5.0)
+        wrapper = RateLimitedClient(client, limiter)
+        resp = await wrapper.get(url)
+    """
+
+    def __init__(self, client, limiter: TokenBucket) -> None:
+        self._client = client
+        self._limiter = limiter
+
+    async def get(self, url: str, **kwargs):
+        """限流的 GET 请求。"""
+        await self._limiter.acquire()
+        return await self._client.get(url, **kwargs)
+
+    async def post(self, url: str, **kwargs):
+        """限流的 POST 请求。"""
+        await self._limiter.acquire()
+        return await self._client.post(url, **kwargs)

src/data/retry_worker.py

@@ -0,0 +1,126 @@
+"""P2-5: IngestFailure 重试 Worker。
+
+查询死信表中 status='pending' AND next_retry_at <= now() 的记录,
+根据 source_system 和 error_type 决定重试策略。
+作为 FastAPI startup 事件注册,在应用启动时执行一次清理。
+"""
+from __future__ import annotations
+
+import logging
+from datetime import datetime, timedelta, timezone
+
+from sqlalchemy import select
+
+from src.db.base import AsyncSessionLocal
+
+logger = logging.getLogger(__name__)
+
+# 最大重试次数上限
+_MAX_RETRIES = 5
+
+
+async def _retry_ingest_failures(db) -> dict:
+    """重试待处理的采集失败记录。
+
+    查询条件: status='pending' AND next_retry_at <= now()
+    重试策略:
+      - retry_count >= MAX_RETRIES → 标记 abandoned
+      - error_type=fetch_error → 退避重试(更新 next_retry_at)
+      - error_type=normalize_error → 规范化失败通常是数据问题,退避重试
+      - error_type=db_error → 数据库问题,退避重试
+
+    Args:
+        db: SQLAlchemy async session
+
+    Returns:
+        统计 dict: {"retried": int, "abandoned": int, "errors": list[str]}
+    """
+    result = {"retried": 0, "abandoned": 0, "errors": []}
+
+    from src.db.models import IngestFailure
+
+    now = datetime.now(timezone.utc)
+
+    # 查询待重试记录
+    stmt = (
+        select(IngestFailure)
+        .where(IngestFailure.status == "pending")
+        .where(IngestFailure.next_retry_at <= now)
+        .order_by(IngestFailure.next_retry_at)
+        .limit(50)  # 每批最多处理 50 条,避免长时间持有事务
+    )
+
+    try:
+        rows = (await db.execute(stmt)).scalars().all()
+    except Exception as e:
+        logger.warning("retry_worker: query failed: %s", e)
+        result["errors"].append(f"query failed: {e}")
+        return result
+
+    if not rows:
+        return result
+
+    logger.info("retry_worker: found %d pending failures to retry", len(rows))
+
+    for failure in rows:
+        try:
+            # 超过最大重试次数 → 放弃
+            if failure.retry_count >= _MAX_RETRIES:
+                failure.status = "abandoned"
+                failure.resolved_at = now
+                await db.flush()
+                result["abandoned"] += 1
+                logger.info(
+                    "retry_worker: abandoned %s/%s after %d retries",
+                    failure.source_system, failure.source_record_id, failure.retry_count,
+                )
+                continue
+
+            # 退避计算: 5min * 2^retry_count,最大 24 小时
+            backoff_minutes = min(5 * (2 ** failure.retry_count), 1440)
+            next_retry = now + timedelta(minutes=backoff_minutes)
+
+            # 更新重试状态
+            failure.retry_count += 1
+            failure.next_retry_at = next_retry
+            failure.status = "retrying"
+            await db.flush()
+            result["retried"] += 1
+
+            logger.info(
+                "retry_worker: scheduled retry %d/%d for %s/%s (next: %s)",
+                failure.retry_count, _MAX_RETRIES,
+                failure.source_system, failure.source_record_id,
+                next_retry.isoformat(),
+            )
+
+        except Exception as e:
+            await db.rollback()
+            error_msg = f"retry {failure.source_system}/{failure.source_record_id}: {e}"
+            result["errors"].append(error_msg)
+            logger.warning("retry_worker: %s", error_msg)
+
+    return result
+
+
+async def run_retry_worker() -> None:
+    """启动入口: 获取 DB session 并执行重试逻辑。
+
+    设计为幂等操作 — 多次运行不会产生副作用(受 next_retry_at 约束)。
+    """
+    logger.info("retry_worker: starting ingest failure retry sweep")
+
+    try:
+        async with AsyncSessionLocal() as db:
+            try:
+                stats = await _retry_ingest_failures(db)
+                await db.commit()
+                logger.info(
+                    "retry_worker: sweep complete — retried=%d abandoned=%d errors=%d",
+                    stats["retried"], stats["abandoned"], len(stats["errors"]),
+                )
+            except Exception as e:
+                await db.rollback()
+                logger.warning("retry_worker: session failed: %s", e)
+    except Exception as e:
+        logger.warning("retry_worker: failed to get DB session: %s", e)

src/data/understat.py

@@ -2,6 +2,12 @@
 
 迁移自旧项目 app/data/sources/understat.py,改成 async。
 使用 Repository 模式进行数据访问,不直接控制事务。
+
+改进(P1):
+  - Bronze 层集成:采集后先存 RawEvent,再规范化
+  - 死信表集成:采集/规范化失败写 IngestFailure
+  - xG 覆盖更新:当 understat 数据更新时覆盖旧值(全量覆盖模式)
+  - xG 追踪字段:xg_source / xg_updated_at / xg_source_record_id
 """
 from __future__ import annotations
 
@@ -10,15 +16,17 @@ import json
 import logging
 import random
 import re
-from datetime import datetime, timezone
+import uuid
+from datetime import datetime, timedelta, timezone
 
-from sqlalchemy import func, select
+from sqlalchemy import select
+from sqlalchemy.orm import selectinload
 
 from src.core.http_client import get_client
 from src.data.config import FDCO_TO_UNDERSTAT, LEAGUE_NAMES
 from src.data.normalize import normalize_understat
 from src.data.sources import register
-from src.db.models import League, Match, MatchStats, Team
+from src.db.models import DataLineage, DataQualityCheck, IngestFailure, League, Match, MatchStats, RawEvent, Team
 
 logger = logging.getLogger(__name__)
 
@@ -46,6 +54,7 @@ async def fetch_understat(league_code: str, season: int) -> list[dict]:
         "Referer": f"https://understat.com/league/{understat_league}/{season}",
     }
 
+
     # 重试:网络错误 / 5xx / 429
     last_exc: Exception | None = None
     for attempt in range(3):
@@ -57,12 +66,10 @@ async def fetch_understat(league_code: str, season: int) -> list[dict]:
         except Exception as e:
             last_exc = e
             if attempt == 2:
-                raise
+                raise RuntimeError(f"understat fetch failed: {e}") from e
             delay = min(2 ** attempt, 8) + random.uniform(0, 1)
             logger.warning("understat fetch failed, retry %d in %.1fs: %s", attempt + 1, delay, e)
             await asyncio.sleep(delay)
-    else:
-        raise RuntimeError(f"understat fetch failed: {last_exc}")
 
     # understat 返回 JS 对象,需要提取 JSON
     text = resp.text
@@ -76,6 +83,148 @@ async def fetch_understat(league_code: str, season: int) -> list[dict]:
     return data
 
 
+def _match_key(home_team_id: int, away_team_id: int, match_date) -> tuple[int, int, str]:
+    """比赛去重键:(主队, 客队, 天级日期 ISO 字符串)。
+
+    统一在这里构造,避免"预加载时用 str(date)、写入时用 isoformat()"这类
+    隐式格式依赖 —— 两者当前恰好相等,但一旦有人改动其一就会静默失配,
+    导致所有比赛被判为不存在而重复插入。
+    """
+    if hasattr(match_date, "date") and callable(match_date.date):
+        match_date = match_date.date()
+    return (home_team_id, away_team_id, match_date.isoformat() if match_date is not None else "")
+
+
+async def _write_understat_bronze(db, raw_matches: list[dict], batch_id: str) -> None:
+    """将 understat 原始数据写入 Bronze 层(RawEvent 表)。
+
+    P2-1: 写入血缘记录,追踪 Bronze 层摄取过程。
+    """
+    for raw in raw_matches:
+        try:
+            source_record_id = str(raw.get("id", ""))
+            if not source_record_id:
+                # P1-fix: 不再静默跳过,记录到死信表
+                logger.warning("understat raw match missing id: %s", str(raw)[:200])
+                await _write_ingest_failure(
+                    db, "understat", "match", "",
+                    "normalize_error", "raw match missing id", raw,
+                )
+                continue
+            bronze = RawEvent(
+                source_system="understat",
+                source_record_id=source_record_id,
+                raw_payload=raw,
+                ingest_batch_id=batch_id,
+            )
+            db.add(bronze)
+        except Exception as e:
+            logger.warning("understat raw_event write failed: %s", e)
+    try:
+        await db.flush()
+    except Exception as e:
+        logger.warning("understat raw_events flush failed: %s", e)
+
+    # P2-1: 写入 Bronze 层血缘记录
+    for raw in raw_matches:
+        source_record_id = str(raw.get("id", ""))
+        if not source_record_id:
+            continue
+        await _write_data_lineage(
+            db,
+            source_system="understat",
+            source_record_id=source_record_id,
+            target_table="raw_events",
+            transform_name="bronze_ingest",
+            batch_id=batch_id,
+        )
+
+
+async def _write_ingest_failure(
+    db,
+    source_system: str,
+    entity_type: str,
+    source_record_id: str,
+    error_type: str,
+    error_detail: str,
+    raw_payload: dict | None,
+) -> None:
+    """写入采集失败到死信表(IngestFailure)。"""
+    try:
+        failure = IngestFailure(
+            source_system=source_system,
+            entity_type=entity_type,
+            source_record_id=source_record_id,
+            error_type=error_type,
+            error_detail=error_detail[:2000] if error_detail else None,
+            raw_payload=raw_payload,
+            status="pending",
+            next_retry_at=datetime.now(timezone.utc) + timedelta(minutes=5),
+        )
+        db.add(failure)
+        await db.flush()
+    except Exception:
+        logger.warning("understat ingest_failure write failed", exc_info=True)
+
+
+async def _write_data_lineage(
+    db,
+    *,
+    source_system: str,
+    source_record_id: str,
+    target_table: str,
+    target_id: int | None = None,
+    transform_name: str,
+    transform_detail: str | None = None,
+    batch_id: str | None = None,
+) -> None:
+    """写入数据血缘记录(DataLineage 表)。"""
+    try:
+        lineage = DataLineage(
+            source_system=source_system,
+            source_record_id=source_record_id,
+            target_table=target_table,
+            target_id=target_id,
+            transform_name=transform_name,
+            transform_detail=transform_detail,
+            batch_id=batch_id,
+        )
+        db.add(lineage)
+        await db.flush()
+    except Exception:
+        logger.warning("understat data_lineage write failed", exc_info=True)
+
+
+async def _write_data_quality_check(
+    db,
+    *,
+    check_name: str,
+    entity_type: str,
+    entity_id: str | None = None,
+    expected_value: str | None = None,
+    actual_value: str | None = None,
+    passed: bool,
+    severity: str = "warning",
+    detail: str | None = None,
+) -> None:
+    """写入数据质量检查结果(DataQualityCheck 表)。"""
+    try:
+        check = DataQualityCheck(
+            check_name=check_name,
+            entity_type=entity_type,
+            entity_id=entity_id,
+            expected_value=expected_value,
+            actual_value=actual_value,
+            passed=passed,
+            severity=severity,
+            detail=detail,
+        )
+        db.add(check)
+        await db.flush()
+    except Exception:
+        logger.warning("understat data_quality_check write failed", exc_info=True)
+
+
 @register
 class UnderstatSource:
     """understat xG 数据源(实现 DataSource 协议)。"""
@@ -83,25 +232,36 @@ class UnderstatSource:
     name = "understat"
 
     async def ingest(self, db, *, league: str, season: int) -> dict:
-        """采集 understat xG → 回填到现有 Match。只回填 xG 字段,不创建新 Match。
+        """采集 understat xG → 回填到现有 Match。
 
         注意: 本方法不控制事务(commit/rollback),由调用方通过 UnitOfWork 控制。
+
+        P1-3: 批量查询优化,将单赛季 380 场 × 3 次 DB 往返降为 3 次查询。
+        P1-4: xG 覆盖更新模式,当 understat 数据更新时覆盖旧值。
         """
-        from src.db.repositories import LeagueRepository, MatchRepository, TeamRepository
+        from src.db.repositories import LeagueRepository, TeamRepository
 
         result = {"updated": 0, "skipped": 0, "unmatched": 0, "errors": []}
+        batch_id = uuid.uuid4().hex[:16]
 
         try:
             raw_matches = await fetch_understat(league, season)
         except Exception as e:
             logger.exception("understat fetch failed for %s %s", league, season)
             result["errors"].append(f"fetch failed: {e}")
+            # 写死信表
+            await _write_ingest_failure(
+                db, "understat", "match", f"fetch_{league}_{season}_{batch_id}",
+                "fetch_error", str(e), {"league": league, "season": season},
+            )
             return result
 
+        # === Bronze 层:先存 RawEvent ===
+        await _write_understat_bronze(db, raw_matches, batch_id)
+
         # 使用 Repository
         league_repo = LeagueRepository(db)
         team_repo = TeamRepository(db)
-        match_repo = MatchRepository(db)
 
         # 查联赛
         league_obj = await league_repo.get_by_code(league)
@@ -109,6 +269,9 @@ class UnderstatSource:
             result["errors"].append(f"league {league} not found in DB")
             return result
 
+        # === 批量优化: 一次规范化,收集球队名和日期 ===
+        normalized_matches: list = []
+        all_team_names: set[str] = set()
         for raw in raw_matches:
             if not raw.get("isResult"):
                 continue
@@ -119,40 +282,126 @@ class UnderstatSource:
                     continue
             except Exception as e:
                 result["errors"].append(f"normalize: {e}")
+                # 写死信表:规范化失败
+                await _write_ingest_failure(
+                    db, "understat", "match", str(raw.get("id", "")),
+                    "normalize_error", str(e), raw,
+                )
                 continue
+            normalized_matches.append((nm, raw))
+            all_team_names.add(nm.home_team)
+            all_team_names.add(nm.away_team)
 
-            # 匹配已有 Match(天级) - 使用 Repository
-            home_team = await team_repo.get_by_name(nm.home_team)
-            away_team = await team_repo.get_by_name(nm.away_team)
-            if home_team is None or away_team is None:
+        if not normalized_matches:
+            return result
+
+        # === 批量查询球队(1 次 DB 往返) ===
+        team_name_to_id = {}
+        if all_team_names:
+            teams = await team_repo.get_all_by_names(list(all_team_names))
+            team_name_to_id = {name: team.id for name, team in teams.items()}
+
+        # === 批量查询已有比赛(1 次 DB 往返,按日期范围) ===
+        match_dict: dict[tuple, Match] = {}
+        dates = [nm.date for nm, _ in normalized_matches if nm.date is not None]
+        if dates:
+            min_dt = min(dates) - timedelta(days=30)
+            max_dt = max(dates) + timedelta(days=30)
+            stmt = (
+                select(Match)
+                .options(selectinload(Match.stats))
+                .where(Match.league_id == league_obj.id)
+                .where(Match.match_date >= min_dt)
+                .where(Match.match_date <= max_dt)
+            )
+            for m in (await db.execute(stmt)).scalars():
+                key = _match_key(m.home_team_id, m.away_team_id, m.match_date_date)
+                match_dict[key] = m
+
+        # === 内存匹配 + 回填 xG(覆盖模式) ===
+        for nm, raw in normalized_matches:
+            home_team_id = team_name_to_id.get(nm.home_team)
+            away_team_id = team_name_to_id.get(nm.away_team)
+            if home_team_id is None or away_team_id is None:
                 result["unmatched"] += 1
                 continue
 
-            existing = await match_repo.find_by_teams_and_date(
-                league_obj.id, home_team.id, away_team.id, nm.date
-            )
+            match_key = _match_key(home_team_id, away_team_id, nm.date)
+            existing = match_dict.get(match_key)
             if existing is None:
                 result["unmatched"] += 1
                 continue
 
-            # 回填 xG
+            now = datetime.now(timezone.utc)
+            source_record_id = str(raw.get("id", ""))
+
+            # 创建 stats 记录(如果不存在)
             if existing.stats is None and (nm.home_xg is not None or nm.away_xg is not None):
-                now = datetime.now(timezone.utc)
                 existing.stats = MatchStats(
                     match_id=existing.id,
                     source="understat",
-                    source_event_id=str(raw.get("id", "")),
+                    source_record_id=source_record_id,
                     retrieved_at=now,
                     available_at=now,
+                    xg_source="understat",
+                    xg_updated_at=now,
+                    xg_source_record_id=source_record_id,
                 )
                 db.add(existing.stats)
                 await db.flush()
+                # P2-1: Silver 层血缘 — MatchStats 创建
+                await _write_data_lineage(
+                    db,
+                    source_system="understat",
+                    source_record_id=source_record_id,
+                    target_table="match_stats",
+                    target_id=existing.id,
+                    transform_name="silver_upsert",
+                    batch_id=batch_id,
+                )
+
+            # xG 覆盖更新模式:当 understat 数据更新时覆盖旧值
             if existing.stats is not None:
-                if existing.stats.home_xg is None and nm.home_xg is not None:
+                xg_updated = False
+                if nm.home_xg is not None:
                     existing.stats.home_xg = nm.home_xg
+                    existing.stats.xg_source = "understat"
+                    existing.stats.xg_updated_at = now
+                    existing.stats.xg_source_record_id = source_record_id
                     result["updated"] += 1
-                if existing.stats.away_xg is None and nm.away_xg is not None:
+                    xg_updated = True
+                if nm.away_xg is not None:
                     existing.stats.away_xg = nm.away_xg
+                    existing.stats.xg_source = "understat"
+                    existing.stats.xg_updated_at = now
+                    existing.stats.xg_source_record_id = source_record_id
+                    xg_updated = True
+                # P2-1: Silver 层血缘 — xG 覆盖更新
+                if xg_updated:
+                    await _write_data_lineage(
+                        db,
+                        source_system="understat",
+                        source_record_id=source_record_id,
+                        target_table="match_stats",
+                        target_id=existing.id,
+                        transform_name="silver_xg_update",
+                        batch_id=batch_id,
+                        transform_detail=f"home_xg={nm.home_xg} away_xg={nm.away_xg}",
+                    )
+
+        # P2-2: 数据质量检查 — 行计数合理性
+        total_processed = result["updated"] + result["skipped"] + result["unmatched"]
+        await _write_data_quality_check(
+            db,
+            check_name="understat_row_count",
+            entity_type="match",
+            entity_id=f"{league}_{season}",
+            expected_value=">=1",
+            actual_value=str(total_processed),
+            passed=total_processed > 0,
+            severity="critical" if total_processed == 0 else "info",
+            detail=f"league={league} season={season} updated={result['updated']} skipped={result['skipped']} unmatched={result['unmatched']}",
+        )
 
         # 注意: 不在此处 commit,由调用方 UnitOfWork 控制事务
         return result

src/db/base.py

@@ -17,8 +17,10 @@ engine = create_async_engine(
     settings.DATABASE_URL,
     echo=False,
     pool_pre_ping=True,
-    pool_size=10,
-    max_overflow=20,
+    pool_size=settings.DB_POOL_SIZE,
+    max_overflow=settings.DB_MAX_OVERFLOW,
+    pool_timeout=settings.DB_POOL_TIMEOUT,
+    pool_recycle=settings.DB_POOL_RECYCLE,
 )
 
 AsyncSessionLocal = async_sessionmaker(

src/db/models.py

@@ -1,4 +1,4 @@
-"""6 张表 ORM: leagues / teams / matches / match_stats / predictions / injuries。"""
+"""ORM 模型: leagues / teams / matches / match_stats / predictions / injuries / raw_events / ingest_failures / data_quality_checks / data_lineage。"""
 from __future__ import annotations
 
 from datetime import date, datetime, timezone
@@ -14,6 +14,7 @@ from sqlalchemy import (
     Integer,
     String,
     Text,
+    UniqueConstraint,
     func,
 )
 from sqlalchemy.dialects.postgresql import JSONB
@@ -75,10 +76,20 @@ class Match(Base):
     created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), default=_utcnow)
     updated_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), default=_utcnow, onupdate=_utcnow)
 
-    league: Mapped[League] = relationship(back_populates="matches")
-    home_team: Mapped[Team] = relationship(foreign_keys=[home_team_id], back_populates="home_matches")
-    away_team: Mapped[Team] = relationship(foreign_keys=[away_team_id], back_populates="away_matches")
-    stats: Mapped["MatchStats | None"] = relationship(back_populates="match", cascade="all, delete-orphan")
+    league: Mapped[League] = relationship(back_populates="matches", lazy="selectin")
+    # lazy="selectin": 这些关系在业务里几乎总是一起读取(切片/回测/展示)。
+    # 默认的 lazy="select" 在 async SQLAlchemy 下,于 session 之外或未显式
+    # eager-load 时访问会抛 MissingGreenlet —— 已因此导致 form/stats/h2h
+    # 三个专家切片静默失败。统一改为预加载,从根上消除这类问题。
+    home_team: Mapped[Team] = relationship(
+        foreign_keys=[home_team_id], back_populates="home_matches", lazy="selectin"
+    )
+    away_team: Mapped[Team] = relationship(
+        foreign_keys=[away_team_id], back_populates="away_matches", lazy="selectin"
+    )
+    stats: Mapped["MatchStats | None"] = relationship(
+        back_populates="match", cascade="all, delete-orphan", lazy="selectin"
+    )
     predictions: Mapped[list["Prediction"]] = relationship(back_populates="match", cascade="all, delete-orphan")
 
     __table_args__ = (
@@ -120,9 +131,20 @@ class MatchStats(Base):
     source_record_id: Mapped[str | None] = mapped_column(String(100))
     retrieved_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True))
     available_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True))
+    # xG 数据追踪:支持 understat 纠正旧 xG 值
+    xg_source: Mapped[str | None] = mapped_column(String(30))  # 具体 xG 数据源
+    xg_updated_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True))  # xG 最后更新时间
+    xg_source_record_id: Mapped[str | None] = mapped_column(String(100))  # xG 对应的源记录 ID
 
     match: Mapped[Match] = relationship(back_populates="stats")
 
+    __table_args__ = (
+        # 数据血缘时间过滤查询用(按 available_at 取「赛前已可得」的统计)
+        Index("ix_match_stats_available_at", "available_at"),
+        # xG 数据源追踪查询用
+        Index("ix_match_stats_xg_source", "xg_source", "xg_updated_at"),
+    )
+
 
 class Injury(Base):
     """球员伤停记录(api-football 数据源)。"""
@@ -183,8 +205,15 @@ class Prediction(Base):
     match: Mapped[Match] = relationship(back_populates="predictions")
 
     __table_args__ = (
+        # P1-6: 数据库级唯一约束,防止同一 match+provider+model 产生重复预测
+        UniqueConstraint(
+            "match_id", "provider", "model",
+            name="uq_predictions_match_provider_model",
+        ),
         Index("ix_predictions_match", "match_id"),
         Index("ix_predictions_provider_model", "provider", "model"),
+        # 数据截止时间过滤查询用(按 prediction_cutoff_at 取「赛前已生成」的预测)
+        Index("ix_predictions_cutoff_at", "prediction_cutoff_at"),
         # 数据库级约束:最后一道防线
         CheckConstraint("pred_home_goals >= 0", name="ck_pred_home_goals_nonneg"),
         CheckConstraint("pred_away_goals >= 0", name="ck_pred_away_goals_nonneg"),
@@ -193,3 +222,108 @@ class Prediction(Base):
         CheckConstraint("mode IN ('single', 'multi')", name="ck_mode_enum"),
         CheckConstraint("status IN ('success', 'failed', 'degraded')", name="ck_status_enum"),
     )
+
+
+# ── Bronze 层:原始事件记录 ──────────────────────────────────────────────
+class RawEvent(Base):
+    """Bronze 层:不可变的原始采集记录。
+
+    每个采集到的原始事件先写入此表,再规范化到 Silver 层(matches / match_stats)。
+    提供完整的数据血缘回溯能力。
+    """
+    __tablename__ = "raw_events"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True)
+    source_system: Mapped[str] = mapped_column(String(30), nullable=False)  # bzzoiro / understat / api-football
+    source_record_id: Mapped[str] = mapped_column(String(100), nullable=False)  # 源系统记录 ID
+    raw_payload: Mapped[dict] = mapped_column(JSONB, nullable=False)  # 完整原始 JSON
+    ingested_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), default=_utcnow, nullable=False)
+    ingest_batch_id: Mapped[str | None] = mapped_column(String(64))  # 批次 ID,用于关联同次采集
+
+    __table_args__ = (
+        UniqueConstraint("source_system", "source_record_id", name="uq_raw_events_source_record"),
+        Index("ix_raw_events_batch", "ingest_batch_id"),
+        Index("ix_raw_events_source_ingested", "source_system", "ingested_at"),
+    )
+
+
+# ── 死信表:采集失败记录 ─────────────────────────────────────────────────
+class IngestFailure(Base):
+    """采集失败死信表。
+
+    当采集或规范化失败时,写入此表而非仅内存 dict。
+    支持自动重试和人工排查。
+    """
+    __tablename__ = "ingest_failures"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True)
+    source_system: Mapped[str] = mapped_column(String(30), nullable=False)
+    entity_type: Mapped[str] = mapped_column(String(30), nullable=False)  # match / injury / team
+    source_record_id: Mapped[str | None] = mapped_column(String(100))
+    error_type: Mapped[str] = mapped_column(String(50), nullable=False)  # fetch_error / normalize_error / db_error / validation_error
+    error_detail: Mapped[str | None] = mapped_column(Text)
+    raw_payload: Mapped[dict | None] = mapped_column(JSONB)  # 失败时的原始数据,用于重试
+    retry_count: Mapped[int] = mapped_column(Integer, default=0)
+    next_retry_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True))
+    status: Mapped[str] = mapped_column(String(20), default="pending")  # pending / retrying / resolved / abandoned
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), default=_utcnow)
+    resolved_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True))
+
+    __table_args__ = (
+        Index("ix_ingest_failures_status_next_retry", "status", "next_retry_at"),
+        Index("ix_ingest_failures_source", "source_system", "entity_type"),
+        CheckConstraint("status IN ('pending', 'retrying', 'resolved', 'abandoned')", name="ck_ingest_failures_status"),
+        CheckConstraint("retry_count >= 0", name="ck_ingest_failures_retry_nonneg"),
+    )
+
+
+# ── 数据质量检查表 ──────────────────────────────────────────────────────
+class DataQualityCheck(Base):
+    """数据质量检查结果记录。
+
+    每次运行数据质量检查时,将结果写入此表用于趋势分析和告警。
+    """
+    __tablename__ = "data_quality_checks"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True)
+    check_name: Mapped[str] = mapped_column(String(100), nullable=False)  # 检查名称
+    entity_type: Mapped[str] = mapped_column(String(30), nullable=False)  # match / team / prediction
+    entity_id: Mapped[str | None] = mapped_column(String(50))  # 具体实体 ID
+    expected_value: Mapped[str | None] = mapped_column(Text)  # 期望值(描述)
+    actual_value: Mapped[str | None] = mapped_column(Text)  # 实际值
+    passed: Mapped[bool] = mapped_column(Boolean, nullable=False)
+    severity: Mapped[str] = mapped_column(String(10), nullable=False, default="warning")  # info / warning / critical
+    detail: Mapped[str | None] = mapped_column(Text)  # 详细描述
+    checked_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), default=_utcnow, nullable=False)
+
+    __table_args__ = (
+        Index("ix_dqc_check_time", "check_name", "checked_at"),
+        Index("ix_dqc_entity", "entity_type", "entity_id"),
+        Index("ix_dqc_severity_passed", "severity", "passed"),
+        CheckConstraint("severity IN ('info', 'warning', 'critical')", name="ck_dqc_severity"),
+    )
+
+
+# ── 数据血缘表 ──────────────────────────────────────────────────────────
+class DataLineage(Base):
+    """ETL 全过程元数据记录。
+
+    追踪从 Bronze → Silver → Gold 的完整转换链路。
+    """
+    __tablename__ = "data_lineage"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True)
+    source_system: Mapped[str] = mapped_column(String(30), nullable=False)
+    source_record_id: Mapped[str] = mapped_column(String(100), nullable=False)
+    target_table: Mapped[str] = mapped_column(String(50), nullable=False)  # matches / match_stats / predictions
+    target_id: Mapped[int | None] = mapped_column(Integer)  # 目标表记录 ID
+    transform_name: Mapped[str] = mapped_column(String(100), nullable=False)  # 转换步骤名称
+    transform_detail: Mapped[str | None] = mapped_column(Text)  # 转换详情
+    batch_id: Mapped[str | None] = mapped_column(String(64))  # 批次 ID
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), default=_utcnow, nullable=False)
+
+    __table_args__ = (
+        Index("ix_lineage_source", "source_system", "source_record_id"),
+        Index("ix_lineage_target", "target_table", "target_id"),
+        Index("ix_lineage_batch", "batch_id"),
+    )

src/db/repositories.py

@@ -5,7 +5,8 @@ Repository 只负责查询,不负责事务提交。
 """
 from __future__ import annotations
 
-from sqlalchemy import func, select
+from datetime import datetime
+from sqlalchemy import select
 from sqlalchemy.orm import selectinload
 from sqlalchemy.ext.asyncio import AsyncSession
 
@@ -37,16 +38,29 @@ class MatchRepository:
     async def find_by_teams_and_date(
         self, league_id: int, home_team_id: int, away_team_id: int, date
     ) -> Match | None:
-        """按联赛+主队+客队+日期查找比赛(天级匹配)。"""
-        if hasattr(date, "date"):
+        """按联赛+主队+客队+日期查找比赛(天级匹配)。
+
+        预加载 stats:调用方(understat 回填)会读取 existing.stats,
+        async session 下惰性加载会抛 MissingGreenlet。
+
+        P2-3: 使用 match_date_date(已建索引)做等值匹配,避免 func.date()
+        导致的全表扫描。
+        """
+        if isinstance(date, datetime):
             date = date.date()
+        elif hasattr(date, "date"):
+            date = date.date()
+        else:
+            # 字符串等其它格式,尝试转换
+            date = datetime.fromisoformat(str(date)).date()
 
         stmt = (
             select(Match)
+            .options(selectinload(Match.stats))
             .where(Match.league_id == league_id)
             .where(Match.home_team_id == home_team_id)
             .where(Match.away_team_id == away_team_id)
-            .where(func.date(Match.match_date) == date)
+            .where(Match.match_date_date == date)
         )
         return (await self._session.execute(stmt)).scalar_one_or_none()
 

src/db/unit_of_work.py

@@ -9,12 +9,16 @@ from __future__ import annotations
 
 from collections.abc import AsyncIterator
 from contextlib import asynccontextmanager
+from typing import TYPE_CHECKING
 
 from src.db.base import AsyncSessionLocal
 
+if TYPE_CHECKING:
+    from sqlalchemy.ext.asyncio import AsyncSession
+
 
 @asynccontextmanager
-async def get_uow() -> AsyncIterator[AsyncSessionLocal]:
+async def get_uow() -> AsyncIterator[AsyncSession]:
     """创建新的工作单元(用于非路由上下文)。
 
     用法:

src/llm/agents/base.py

@@ -12,7 +12,7 @@ import logging
 from dataclasses import dataclass, field
 from pathlib import Path
 
-from src.llm.context_builder import MatchHeader
+from src.llm.context_builder import MatchHeader, SliceResult
 from src.llm.provider import LLMProvider, LLMResponse
 
 logger = logging.getLogger(__name__)
@@ -80,7 +80,12 @@ class AgentReport:
 
 
 def _is_no_data(slice_text: str) -> bool:
-    """切片是否全无数据(除了标题行全是无数据)。"""
+    """兜底: 判断纯字符串切片是否全无数据。
+
+    仅用于 slice_fn 返回 `str`(未升级为 SliceResult)的场景。
+    新代码应让切片返回 SliceResult 并显式声明 has_data —— 字符串子串匹配
+    依赖具体文案(「无比分数据」「无伤停数据」等变体会漏判),不可靠。
+    """
     body = [ln.strip() for ln in slice_text.splitlines() if ln.strip()]
     # 去掉标题行(── 开头)
     content = [ln for ln in body if not ln.startswith("──")]
@@ -89,6 +94,18 @@ def _is_no_data(slice_text: str) -> bool:
     return all(any(s in ln for s in NO_DATA_SENTINELS) for ln in content)
 
 
+def _slice_has_data(slice_result) -> tuple[str, bool]:
+    """把切片返回值统一成 (text, has_data)。
+
+    优先使用 SliceResult.has_data(结构化,可信);若切片函数仍返回 str,
+    则回退到文案子串匹配(向后兼容)。
+    """
+    if isinstance(slice_result, SliceResult):
+        return slice_result.text, slice_result.has_data
+    text = str(slice_result)
+    return text, not _is_no_data(text)
+
+
 def _stub_no_data(agent: str) -> AgentReport:
     return AgentReport(
         agent=agent,
@@ -145,13 +162,14 @@ async def run_agent(
     """执行单个专家 agent: 切片 → no_data 门控 → 调 LLM → 解析报告。"""
     # 1. 数据切片
     try:
-        slice_text = await spec.slice_fn(header, before=before)
+        slice_result = await spec.slice_fn(header, before=before)
     except Exception as e:
         logger.exception("agent %s slice failed", spec.name)
         return AgentReport(agent=spec.name, status="error", analysis=f"数据切片失败: {e}")
 
-    # 2. no_data 门控: 切片无数据 → 不调 LLM
-    if _is_no_data(slice_text):
+    # 2. no_data 门控: 切片显式声明无数据 → 不调 LLM
+    slice_text, has_data = _slice_has_data(slice_result)
+    if not has_data:
         logger.debug("agent %s: slice is no_data, skipping LLM", spec.name)
         return _stub_no_data(spec.name)
 

src/llm/agents/orchestrator.py

@@ -195,13 +195,14 @@ async def predict_match_multi(
             raise ValueError(f"match {match_id} not found")
 
         # 严格校验终裁输出
-        from src.llm.validation import validate_prediction_output
+        from src.llm.validation import validate_agent_weights, validate_prediction_output
         try:
             validated = validate_prediction_output(final)
         except Exception as e:
             raise RuntimeError(f"终裁输出校验失败: {e}")
 
-        agent_weights = final.get("agent_weights")
+        # agent_weights 同样必须过校验(旧实现直接取 raw 值落库,未做任何检查)
+        agent_weights = validate_agent_weights(final.get("agent_weights"))
         pred = Prediction(
             match_id=match_id,
             provider=settings.LLM_PROVIDER,

src/llm/backtest.py

@@ -3,18 +3,23 @@
 核心机制:
   - build_context 已内置 before=match_date,天然防未来信息泄漏
   - 对历史比赛跑预测 → 用实际比分 settle → 统计准确率
+  - 并发控制: asyncio.Semaphore 限制同时 LLM 调用数
 """
 from __future__ import annotations
 
+import asyncio
 import logging
 from dataclasses import dataclass, field
+from datetime import datetime
 
-from sqlalchemy import and_, select
+from sqlalchemy import select
+from sqlalchemy.orm import selectinload
 
-from src.db.models import League, Match
+from src.db.models import Match
 from src.db.unit_of_work import get_uow
 from src.llm.eval import settle_prediction
 from src.llm.predict import predict_match
+from src.llm.utils import actual_1x2
 
 logger = logging.getLogger(__name__)
 
@@ -38,6 +43,23 @@ class BacktestMatchResult:
     prediction_id: int
 
 
+@dataclass
+class BacktestCandidate:
+    """回测候选比赛(字段快照,不持有 ORM 对象)。
+
+    session 关闭后仍可安全读取:所有需要的关系字段已在查询时物化为普通值,
+    避免在 session 之外访问惰性加载的关系属性(会抛 MissingGreenlet)。
+    """
+    match_id: int
+    league_code: str | None
+    home_team: str
+    away_team: str
+    match_date: datetime
+    home_goals: int
+    away_goals: int
+
+
+
 @dataclass
 class BacktestSummary:
     """回测汇总统计。"""
@@ -50,15 +72,6 @@ class BacktestSummary:
     results: list[BacktestMatchResult] = field(default_factory=list)
 
 
-def _actual_1x2(home: int, away: int) -> str:
-    """实际比分 → 胜平负。"""
-    if home > away:
-        return "1"
-    if home == away:
-        return "X"
-    return "2"
-
-
 async def _get_historical_matches(
     db,
     *,
@@ -66,10 +79,20 @@ async def _get_historical_matches(
     date_from: str | None = None,
     date_to: str | None = None,
     limit: int = 50,
-) -> list[Match]:
-    """查询已完赛且有比分的比赛(回测候选)。"""
+) -> list[BacktestCandidate]:
+    """查询已完赛且有比分的比赛(回测候选)。
+
+    返回普通值快照而非 ORM 对象:调用方在 session 关闭后仍需使用这些字段,
+    而 league / home_team / away_team 是惰性加载关系,在 async 下于 session
+    之外访问会抛 MissingGreenlet。这里用 selectinload 预加载后立即物化。
+    """
     stmt = (
         select(Match)
+        .options(
+            selectinload(Match.league),
+            selectinload(Match.home_team),
+            selectinload(Match.away_team),
+        )
         .where(Match.match_status == "finished")
         .where(Match.home_goals.is_not(None))
         .where(Match.away_goals.is_not(None))
@@ -83,7 +106,19 @@ async def _get_historical_matches(
 
     stmt = stmt.order_by(Match.match_date.desc()).limit(limit)
     result = await db.execute(stmt)
-    return list(result.scalars().all())
+    # 在 session 内物化为纯数据,切断与 ORM 会话的耦合
+    return [
+        BacktestCandidate(
+            match_id=m.id,
+            league_code=m.league.code if m.league else None,
+            home_team=m.home_team.name if m.home_team else "?",
+            away_team=m.away_team.name if m.away_team else "?",
+            match_date=m.match_date,
+            home_goals=m.home_goals,
+            away_goals=m.away_goals,
+        )
+        for m in result.scalars().all()
+    ]
 
 
 async def run_backtest(
@@ -109,46 +144,48 @@ async def run_backtest(
         BacktestSummary 含逐场结果 + 汇总统计
     """
     async with get_uow() as session:
-        matches = await _get_historical_matches(
+        candidates = await _get_historical_matches(
             session, league_id=league_id, date_from=date_from, date_to=date_to, limit=limit
         )
 
-    summary = BacktestSummary(total=len(matches), scored=0)
+    summary = BacktestSummary(total=len(candidates), scored=0)
 
-    for m in matches:
-        try:
-            # 预测 (build_context 内部已用 before=match_date 防泄漏,
-            # injuries_slice 也使用 as_of=match_date 过滤 retrieved_at)
-            result = await predict_match(m.id, mode=mode, model=model)
+    # P2: 并发控制,同时最多 3 场预测(避免 LLM API 限流,保护下游服务)
+    sem = asyncio.Semaphore(3)
 
-            # 用实际比分 settle
-            await settle_prediction(result.prediction_id, m.home_goals, m.away_goals)
+    async def _one(c: BacktestCandidate) -> BacktestMatchResult | None:
+        async with sem:
+            try:
+                result = await predict_match(c.match_id, mode=mode, model=model, use_cache=False, backtest=True)
+                await settle_prediction(result.prediction_id, c.home_goals, c.away_goals)
+                actual = actual_1x2(c.home_goals, c.away_goals)
+                return BacktestMatchResult(
+                    match_id=c.match_id,
+                    league_code=c.league_code,
+                    home_team=c.home_team,
+                    away_team=c.away_team,
+                    match_date=c.match_date.strftime("%Y-%m-%d") if c.match_date else "?",
+                    actual_home=c.home_goals,
+                    actual_away=c.away_goals,
+                    actual_1x2=actual,
+                    pred_home=result.pred_home_goals,
+                    pred_away=result.pred_away_goals,
+                    pred_1x2=result.pred_1x2,
+                    subjective_confidence=result.subjective_confidence,
+                    correct_1x2=result.pred_1x2 == actual,
+                    prediction_id=result.prediction_id,
+                )
+            except Exception:
+                logger.exception("backtest match %s failed", c.match_id)
+                return None
 
-            actual = _actual_1x2(m.home_goals, m.away_goals)
-            correct = result.pred_1x2 == actual
-
-            bt = BacktestMatchResult(
-                match_id=m.id,
-                league_code=m.league.code if m.league else None,
-                home_team=m.home_team.name if m.home_team else "?",
-                away_team=m.away_team.name if m.away_team else "?",
-                match_date=m.match_date.strftime("%Y-%m-%d") if m.match_date else "?",
-                actual_home=m.home_goals,
-                actual_away=m.away_goals,
-                actual_1x2=actual,
-                pred_home=result.pred_home_goals,
-                pred_away=result.pred_away_goals,
-                pred_1x2=result.pred_1x2,
-                subjective_confidence=result.subjective_confidence,
-                correct_1x2=correct,
-                prediction_id=result.prediction_id,
-            )
-            summary.results.append(bt)
+    # 并行执行,保持结果顺序
+    results = await asyncio.gather(*[_one(c) for c in candidates])
+    for r in results:
+        if r is not None:
+            summary.results.append(r)
             summary.scored += 1
 
-        except Exception as e:
-            logger.warning("backtest match %s failed: %s", m.id, e)
-
     # 汇总统计
     if summary.scored > 0:
         correct_count = sum(1 for r in summary.results if r.correct_1x2)

src/llm/context_builder.py

@@ -1,16 +1,22 @@
 """上下文构建器:数据切片 + 拼接。
 
 架构:
-  - match_header: 比赛基础信息(对阵双方/联赛/时间)
-  - 切片函数: 每个领域 agent 一个数据切片(h2h / form / standings / injuries / xg)
-  - build_context: 单 agent 路径,拼接全部切片(行为与旧版一致)
+    - match_header: 比赛基础信息(对阵双方/联赛/时间)
+    - 切片函数: 每个领域 agent 一个数据切片(h2h / form / standings / injuries / xg)
+    - build_context: 单 agent 路径,拼接全部切片(行为与旧版一致)
 
 multi-agent 路径由 agents/orchestrator.py 调用切片函数,每个专家只拿自己的切片。
+
+性能说明:
+    build_context 创建一个共享 session 并传给所有切片函数,
+    避免每个切片独立创建 session —— 回测 20 场并发时,
+    5 个切片 × 20 场 = 100 个连接会耗尽连接池(pool_size=15)。
 """
 from __future__ import annotations
 
 import logging
 from dataclasses import dataclass
+from typing import TYPE_CHECKING
 
 from sqlalchemy import select
 from sqlalchemy.orm import selectinload
@@ -18,6 +24,9 @@ from sqlalchemy.orm import selectinload
 from src.db.base import AsyncSessionLocal
 from src.db.models import Match
 
+if TYPE_CHECKING:
+    from sqlalchemy.ext.asyncio import AsyncSession
+
 logger = logging.getLogger(__name__)
 
 
@@ -39,6 +48,22 @@ def _is_stats_available(stats, before) -> bool:
     return stats.available_at <= before
 
 
+@dataclass
+class SliceResult:
+    """数据切片的显式结果(替代「靠文案子串猜有无数据」)。
+
+    旧实现用 `"无数据" in slice_text` 判断,依赖具体文案 —— 一旦某个切片
+    写成「无比分数据」「无伤停数据」这类变体,判断就会静默失配
+    (见审查报告 P2-1)。这里让切片函数直接声明 `has_data`,不再猜。
+    """
+    text: str
+    has_data: bool
+    n_records: int = 0
+
+    def __str__(self) -> str:  # 让老调用点可直接当 str 用
+        return self.text
+
+
 @dataclass
 class MatchContext:
     match_id: int
@@ -64,11 +89,19 @@ class MatchHeader:
     league_id: int
 
 
-async def load_match_header(match_id: int) -> MatchHeader:
-    """加载比赛头信息(各 agent 共用)。"""
-    async with AsyncSessionLocal() as db:
+async def load_match_header(match_id: int, db: AsyncSession | None = None) -> MatchHeader:
+    """加载比赛头信息(各 agent 共用)。
+
+    Args:
+        match_id: 比赛 ID
+        db: 可选的共享 session。不传则自建(向后兼容)。
+    """
+    if db is not None:
         m = await _load_match(db, match_id)
         return _to_header(m)
+    async with AsyncSessionLocal() as new_db:
+        m = await _load_match(new_db, match_id)
+        return _to_header(m)
 
 
 def _to_header(m: Match) -> MatchHeader:
@@ -98,16 +131,24 @@ def header_text(h: MatchHeader) -> str:
 # 切片函数: 每个领域 agent 一个
 # ============================================================
 
-async def h2h_slice(header: MatchHeader, *, limit: int = 8, before=None) -> str:
-    """E - 历史交锋切片: 过去数年 + 近期交手数据,提取交手规律。before=match_date 用于回测。"""
-    async with AsyncSessionLocal() as db:
+async def h2h_slice(header: MatchHeader, *, limit: int = 8, before=None, db: AsyncSession | None = None) -> SliceResult:
+    """E - 历史交锋切片: 过去数年 + 近期交手数据,提取交手规律。before=match_date 用于回测。
+
+    db: 可选共享 session,避免每个切片独立建连(见模块 docstring)。
+    """
+    if db is not None:
         h2h = await _get_h2h(db, header.home_team_id, header.away_team_id, before=before, limit=limit)
+    else:
+        async with AsyncSessionLocal() as new_db:
+            h2h = await _get_h2h(new_db, header.home_team_id, header.away_team_id, before=before, limit=limit)
     lines = [f"── 历史交锋(近 {limit} 次) ──"]
+    n_with_score = 0
     if h2h:
         home_wins = draws = away_wins = 0
         for hm in h2h:
             d = hm.match_date.strftime("%Y-%m") if hm.match_date else "?"
             if hm.home_goals is not None:
+                n_with_score += 1
                 if hm.home_goals > hm.away_goals: home_wins += 1
                 elif hm.home_goals == hm.away_goals: draws += 1
                 else: away_wins += 1
@@ -119,15 +160,24 @@ async def h2h_slice(header: MatchHeader, *, limit: int = 8, before=None) -> str:
             lines.append(f"  总计 {total} 场: 主队 {home_wins}胜 {draws}平 {away_wins}负")
     else:
         lines.append("  无数据")
-    return "\n".join(lines)
+    # has_data 以「有比分的交锋」为准:仅有对阵无比分时不足以支撑分析
+    return SliceResult(text="\n".join(lines), has_data=n_with_score > 0, n_records=n_with_score)
 
 
-async def form_slice(header: MatchHeader, *, limit: int = 5, before=None) -> str:
-    """A - 近期状态切片: 两队近 N 场赛果、关键事件、走势判断。before=match_date 用于回测。"""
-    async with AsyncSessionLocal() as db:
+async def form_slice(header: MatchHeader, *, limit: int = 5, before=None, db: AsyncSession | None = None) -> SliceResult:
+    """A - 近期状态切片: 两队近 N 场赛果、关键事件、走势判断。before=match_date 用于回测。
+
+    db: 可选共享 session,避免每个切片独立建连(见模块 docstring)。
+    """
+    if db is not None:
         home_form = await _get_form(db, header.home_team_id, before=before, limit=limit)
         away_form = await _get_form(db, header.away_team_id, before=before, limit=limit)
+    else:
+        async with AsyncSessionLocal() as new_db:
+            home_form = await _get_form(new_db, header.home_team_id, before=before, limit=limit)
+            away_form = await _get_form(new_db, header.away_team_id, before=before, limit=limit)
     lines = []
+    n_scored = 0
     for label, name, form, side in (
         ("主队", header.home_name, home_form, "home"),
         ("客队", header.away_name, away_form, "away"),
@@ -140,6 +190,8 @@ async def form_slice(header: MatchHeader, *, limit: int = 5, before=None) -> str
                 if o == "W": wins += 1
                 elif o == "D": draws += 1
                 else: losses += 1
+                if fm.home_goals is not None:
+                    n_scored += 1
                 score = f"{fm.home_goals}-{fm.away_goals}" if fm.home_goals is not None else "vs"
                 xg = ""
                 if fm.stats and _is_stats_available(fm.stats, before) and fm.stats.home_xg is not None:
@@ -150,15 +202,23 @@ async def form_slice(header: MatchHeader, *, limit: int = 5, before=None) -> str
             lines.append(f"  近 {len(form)} 场: {wins}胜 {draws}平 {losses}负")
         else:
             lines.append("  无数据")
-    return "\n".join(lines)
+    return SliceResult(text="\n".join(lines), has_data=n_scored > 0, n_records=n_scored)
 
 
-async def stats_slice(header: MatchHeader, *, limit: int = 10, before=None) -> str:
-    """B - 攻防数据切片: 进球、射门、控球,评估攻防强度。before=match_date 用于回测。"""
-    async with AsyncSessionLocal() as db:
+async def stats_slice(header: MatchHeader, *, limit: int = 10, before=None, db: AsyncSession | None = None) -> SliceResult:
+    """B - 攻防数据切片: 进球、射门、控球,评估攻防强度。before=match_date 用于回测。
+
+    db: 可选共享 session,避免每个切片独立建连(见模块 docstring)。
+    """
+    if db is not None:
         home_form = await _get_form(db, header.home_team_id, before=before, limit=limit)
         away_form = await _get_form(db, header.away_team_id, before=before, limit=limit)
+    else:
+        async with AsyncSessionLocal() as new_db:
+            home_form = await _get_form(new_db, header.home_team_id, before=before, limit=limit)
+            away_form = await _get_form(new_db, header.away_team_id, before=before, limit=limit)
     lines = [f"── 攻防数据(近 {limit} 场) ──"]
+    n_total = 0
     for label, name, form, side in (
         ("主队", header.home_name, home_form, "home"),
         ("客队", header.away_name, away_form, "away"),
@@ -184,6 +244,7 @@ async def stats_slice(header: MatchHeader, *, limit: int = 10, before=None) -> s
                         xg += fm.stats.home_xg if side == "home" else fm.stats.away_xg
                         xga += fm.stats.away_xg if side == "home" else fm.stats.home_xg
                         n_xg += 1
+            n_total += n
             if n > 0:
                 lines.append(f"  {label} {name}:")
                 lines.append(f"    场均进球 {gf/n:.2f}, 场均失球 {ga/n:.2f}")
@@ -194,15 +255,23 @@ async def stats_slice(header: MatchHeader, *, limit: int = 10, before=None) -> s
                 lines.append(f"  {label} {name}: 无比分数据")
         else:
             lines.append(f"  {label} {name}: 无数据")
-    return "\n".join(lines)
+    return SliceResult(text="\n".join(lines), has_data=n_total > 0, n_records=n_total)
 
 
-async def home_away_slice(header: MatchHeader, *, limit: int = 10, before=None) -> str:
-    """C - 主客因素切片: 主场战绩 vs 客场战绩,评估地理优势影响。before=match_date 用于回测。"""
-    async with AsyncSessionLocal() as db:
+async def home_away_slice(header: MatchHeader, *, limit: int = 10, before=None, db: AsyncSession | None = None) -> SliceResult:
+    """C - 主客因素切片: 主场战绩 vs 客场战绩,评估地理优势影响。before=match_date 用于回测。
+
+    db: 可选共享 session,避免每个切片独立建连(见模块 docstring)。
+    """
+    if db is not None:
         home_home = await _get_home_away(db, header.home_team_id, "home", before=before, limit=limit)
         away_away = await _get_home_away(db, header.away_team_id, "away", before=before, limit=limit)
+    else:
+        async with AsyncSessionLocal() as new_db:
+            home_home = await _get_home_away(new_db, header.home_team_id, "home", before=before, limit=limit)
+            away_away = await _get_home_away(new_db, header.away_team_id, "away", before=before, limit=limit)
     lines = ["── 主客因素 ──"]
+    n_total = 0
     for label, name, matches, side in (
         ("主队主场", header.home_name, home_home, "home"),
         ("客队客场", header.away_name, away_away, "away"),
@@ -218,6 +287,7 @@ async def home_away_slice(header: MatchHeader, *, limit: int = 10, before=None)
                 gf += m.home_goals if side == "home" else m.away_goals
                 ga += m.away_goals if side == "home" else m.home_goals
             n = wins + draws + losses
+            n_total += n
             if n > 0:
                 pct = wins / n * 100
                 lines.append(f"  {label} {name}(近 {n} 场): {wins}胜 {draws}平 {losses}负, 胜率 {pct:.0f}%")
@@ -226,26 +296,31 @@ async def home_away_slice(header: MatchHeader, *, limit: int = 10, before=None)
                 lines.append(f"  {label} {name}: 无比分数据")
         else:
             lines.append(f"  {label} {name}: 无数据")
-    return "\n".join(lines)
+    return SliceResult(text="\n".join(lines), has_data=n_total > 0, n_records=n_total)
 
 
-async def injuries_slice(header: MatchHeader, *, before=None) -> str:
+async def injuries_slice(header: MatchHeader, *, before=None, db: AsyncSession | None = None) -> SliceResult:
     """D - 阵容完整性切片: 伤停与停赛名单,评估战力缺失程度。
 
     before=cutoff: 只使用 cutoff 之前已采集的伤停数据,防回测泄漏。
+    db: 可选共享 session(见模块 docstring)。
     """
     from src.data.injuries import get_injuries_for_match
 
     cutoff = before or header.match_dt
-    async with AsyncSessionLocal() as db:
+    if db is not None:
         home_injuries = await get_injuries_for_match(db, header.home_team_id, cutoff, as_of=cutoff)
         away_injuries = await get_injuries_for_match(db, header.away_team_id, cutoff, as_of=cutoff)
+    else:
+        async with AsyncSessionLocal() as new_db:
+            home_injuries = await get_injuries_for_match(new_db, header.home_team_id, cutoff, as_of=cutoff)
+            away_injuries = await get_injuries_for_match(new_db, header.away_team_id, cutoff, as_of=cutoff)
 
     lines = ["── 阵容完整性 ──"]
-    has_data = False
+    n_records = 0
     for label, injuries in (("主队", home_injuries), ("客队", away_injuries)):
         if injuries:
-            has_data = True
+            n_records += len(injuries)
             lines.append(f"  {label}伤停({len(injuries)}人):")
             for inj in injuries[:8]:  # 最多显示 8 条
                 reason = inj.reason or inj.injury_type or "未知"
@@ -255,55 +330,61 @@ async def injuries_slice(header: MatchHeader, *, before=None) -> str:
         else:
             lines.append(f"  {label}: 无伤停数据")
 
-    if not has_data:
-        return "── 阵容完整性 ──\n  无数据"
+    if n_records == 0:
+        return SliceResult(text="── 阵容完整性 ──\n  无数据", has_data=False, n_records=0)
 
-    return "\n".join(lines)
+    return SliceResult(text="\n".join(lines), has_data=True, n_records=n_records)
 
 
 # ============================================================
 # 单 agent 路径: 拼接全部切片(行为与旧版一致)
 # ============================================================
 
-async def build_context(match_id: int, *, form_last: int = 5, h2h_last: int = 5) -> MatchContext:
-    """单 agent 路径的完整上下文: 拼接全部切片(before=比赛时间,防未来信息)。"""
-    header = await load_match_header(match_id)
-    parts = [header_text(header), ""]
-    has_stats = False
-    has_injuries = False
+async def build_context(match_id: int, *, form_last: int = 5, h2h_last: int = 5, backtest: bool = False) -> MatchContext:
+    """单 agent 路径的完整上下文: 拼接全部切片(before=比赛时间,防未来信息)。
 
-    form_text = await form_slice(header, limit=form_last, before=header.match_dt)
-    if "无数据" not in form_text:
-        has_stats = True
-    parts.append(form_text)
-    parts.append("")
+    has_stats / has_injuries 直接取切片显式声明的 has_data,
+    不再靠文案子串匹配(见审查报告 P2-1)。
 
-    h2h_text = await h2h_slice(header, limit=h2h_last, before=header.match_dt)
-    parts.append(h2h_text)
-    parts.append("")
+    P2-6: backtest=True 时 cutoff = match_date - 1天,确保只用赛前数据。
 
-    stats_text = await stats_slice(header, before=header.match_dt)
-    if "无数据" not in stats_text:
-        has_stats = True
-    parts.append(stats_text)
-    parts.append("")
+    P1-1: 使用单个共享 session 贯穿所有切片查询,避免连接池耗尽。
+    """
+    async with AsyncSessionLocal() as db:
+        header = await load_match_header(match_id, db=db)
+        # P2-6: 回测模式下 cutoff 提前 1 天,防止比赛日数据泄漏
+        cutoff = header.match_dt
+        if backtest and header.match_dt:
+            from datetime import timedelta
+            cutoff = header.match_dt - timedelta(days=1)
+        parts = [header_text(header), ""]
 
-    home_away_text = await home_away_slice(header, before=header.match_dt)
-    parts.append(home_away_text)
-    parts.append("")
+        form_res = await form_slice(header, limit=form_last, before=cutoff, db=db)
+        parts.append(form_res.text)
+        parts.append("")
 
-    injuries_text = await injuries_slice(header, before=header.match_dt)
-    if "无数据" not in injuries_text:
-        has_injuries = True
-    parts.append(injuries_text)
+        h2h_res = await h2h_slice(header, limit=h2h_last, before=cutoff, db=db)
+        parts.append(h2h_res.text)
+        parts.append("")
 
-    return MatchContext(
-        match_id=match_id,
-        text="\n".join(parts),
-        has_stats=has_stats,
-        has_injuries=has_injuries,
-        match_dt=header.match_dt,
-    )
+        stats_res = await stats_slice(header, before=cutoff, db=db)
+        parts.append(stats_res.text)
+        parts.append("")
+
+        home_away_res = await home_away_slice(header, before=cutoff, db=db)
+        parts.append(home_away_res.text)
+        parts.append("")
+
+        injuries_res = await injuries_slice(header, before=cutoff, db=db)
+        parts.append(injuries_res.text)
+
+        return MatchContext(
+            match_id=match_id,
+            text="\n".join(parts),
+            has_stats=form_res.has_data or stats_res.has_data,
+            has_injuries=injuries_res.has_data,
+            match_dt=header.match_dt,
+        )
 
 
 # ============================================================
@@ -328,9 +409,19 @@ async def _load_match(db, match_id: int) -> Match:
 
 
 async def _get_form(db, team_id: int, before, *, limit: int = 5) -> list[Match]:
-    """某队近 N 场(已完赛)。before=None 表示不限制(预测赛前的场景由调用方保证)。"""
+    """某队近 N 场(已完赛)。before=None 表示不限制(预测赛前的场景由调用方保证)。
+
+    必须预加载 stats / home_team / away_team:切片函数会读取这些关系,
+    而 async session 下惰性加载会抛 MissingGreenlet。
+    (models.py 已声明 lazy="selectin",此处显式声明以固化查询意图。)
+    """
     stmt = (
         select(Match)
+        .options(
+            selectinload(Match.stats),
+            selectinload(Match.home_team),
+            selectinload(Match.away_team),
+        )
         .where(Match.match_status == "finished")
         .where(Match.home_goals.is_not(None))
         .where((Match.home_team_id == team_id) | (Match.away_team_id == team_id))
@@ -344,9 +435,13 @@ async def _get_form(db, team_id: int, before, *, limit: int = 5) -> list[Match]:
 
 
 async def _get_h2h(db, home_id: int, away_id: int, before, *, limit: int = 5) -> list[Match]:
-    """两队交锋史。"""
+    """两队交锋史。需预加载 home_team / away_team(切片输出队名)。"""
     stmt = (
         select(Match)
+        .options(
+            selectinload(Match.home_team),
+            selectinload(Match.away_team),
+        )
         .where(Match.match_status == "finished")
         .where(Match.home_goals.is_not(None))
         .where(
@@ -363,9 +458,17 @@ async def _get_h2h(db, home_id: int, away_id: int, before, *, limit: int = 5) ->
 
 
 async def _get_home_away(db, team_id: int, side: str, before, *, limit: int = 10) -> list[Match]:
-    """某队主场/客场近 N 场。side='home' 取主场,'away' 取客场。"""
+    """某队主场/客场近 N 场。side='home' 取主场,'away' 取客场。
+
+    当前只用标量字段,但统一预加载以免后续扩展时踩坑。
+    """
     stmt = (
         select(Match)
+        .options(
+            selectinload(Match.stats),
+            selectinload(Match.home_team),
+            selectinload(Match.away_team),
+        )
         .where(Match.match_status == "finished")
         .where(Match.home_goals.is_not(None))
         .order_by(Match.match_date.desc())

src/llm/predict.py

@@ -23,29 +23,47 @@ _PROMPT_DIR = Path(__file__).resolve().parent / "prompts"
 
 # ── LLM 响应缓存(match+provider+model+version → 结果) ──
 _CACHE_TTL_SEC = 300  # 5 分钟
+# P1-5: 缓存仅在 asyncio 协程内同步访问(dict 操作 GIL 原子),无需 threading.Lock。
+# 删除 _cache_lock,避免同步锁阻塞事件循环;dict 的 get/set 在 CPython 下原子。
 _cache: dict[str, tuple[float, PredictResult]] = {}
-_cache_lock = Lock()
 
 
-def _cache_key(match_id: int, provider: str, model: str, version: str) -> str:
-    return f"{match_id}:{provider}:{model}:{version}"
+def _cache_key(match_id: int, provider: str, model: str, version: str, tpl_hash: str) -> str:
+    """缓存键:含 prompt 模板内容 hash。
+
+    仅用 version 做键不够 —— 编辑器里改动 `match_prediction_v1.md` 而版本号
+    不变时,进程内缓存仍会返回旧模板产生的旧结果(见审查报告 P2-6)。
+    把模板内容 hash 纳入键,模板一改缓存自动失效。
+    """
+    return f"{match_id}:{provider}:{model}:{version}:{tpl_hash[:12]}"
 
 
-def _get_cached(match_id: int, provider: str, model: str, version: str) -> PredictResult | None:
-    key = _cache_key(match_id, provider, model, version)
-    with _cache_lock:
-        if key in _cache:
-            ts, result = _cache[key]
-            if time.time() - ts < _CACHE_TTL_SEC:
-                return result
-            del _cache[key]
+def _get_cached(match_id: int, provider: str, model: str, version: str, tpl_hash: str) -> PredictResult | None:
+    # P1-5: 无锁访问。dict get/del 在 CPython GIL 下原子,且无 await 穿插。
+    key = _cache_key(match_id, provider, model, version, tpl_hash)
+    entry = _cache.get(key)
+    if entry is not None:
+        ts, result = entry
+        if time.time() - ts < _CACHE_TTL_SEC:
+            return result
+        _cache.pop(key, None)
     return None
 
 
-def _set_cached(match_id: int, provider: str, model: str, version: str, result: PredictResult) -> None:
-    key = _cache_key(match_id, provider, model, version)
-    with _cache_lock:
-        _cache[key] = (time.time(), result)
+def _set_cached(match_id: int, provider: str, model: str, version: str, tpl_hash: str, result: PredictResult) -> None:
+    # P1-5: 无锁写入。同上,dict set 原子。
+    key = _cache_key(match_id, provider, model, version, tpl_hash)
+    _cache[key] = (time.time(), result)
+
+
+def clear_prompt_cache() -> None:
+    """清空 prompt 模板缓存(供开发/热更新时手动调用)。
+
+    lru_cache 的模板缓存是进程级的,改完 .md 需要重启进程才能生效;
+    提供显式清理入口,避免"改了模板却看不到变化"的困惑(见审查报告 P2-5)。
+    """
+    _load_prompt_template.cache_clear()
+    logger.info("prompt 模板缓存已清空")
 
 
 @functools.lru_cache(maxsize=8)
@@ -58,6 +76,11 @@ def _load_prompt_template(version: str = "v1") -> str:
         return f.read()
 
 
+def _prompt_template_hash(version: str) -> str:
+    """prompt 模板内容 hash(用于缓存键,模板变更即失效)。"""
+    return hashlib.sha256(_load_prompt_template(version).encode("utf-8")).hexdigest()
+
+
 @dataclass
 class PredictResult:
     prediction_id: int
@@ -81,11 +104,25 @@ async def predict_match(
     model: str | None = None,
     prompt_version: str | None = None,
     mode: str = "multi",
+    use_cache: bool = True,
+    backtest: bool = False,
 ) -> "PredictResult | MultiPredictResult":
-    """预测入口。mode=multi(默认)走多 agent;mode=single 走单次调用。"""
+    """预测入口。mode=multi(默认)走多 agent;mode=single 走单次调用。
+
+    Args:
+        use_cache: 是否允许返回进程内缓存结果。回测必须传 False——
+            缓存命中不会新建 prediction 行,调用方会对同一个 prediction_id
+            反复 settle,把不同比赛的真实比分覆盖到同一条记录上。
+        backtest: 是否回测模式。True 时 build_context 使用 match_date-1天 作为 cutoff。
+    """
     if mode == "single":
         return await _predict_single(
-            match_id, provider=provider, model=model, prompt_version=prompt_version
+            match_id,
+            provider=provider,
+            model=model,
+            prompt_version=prompt_version,
+            use_cache=use_cache,
+            backtest=backtest,
         )
     from src.llm.agents.orchestrator import predict_match_multi
 
@@ -98,6 +135,8 @@ async def _predict_single(
     provider: LLMProvider | None = None,
     model: str | None = None,
     prompt_version: str | None = None,
+    use_cache: bool = True,
+    backtest: bool = False,
 ) -> PredictResult:
     """单次调用路径(原有实现)。"""
     if provider is None:
@@ -105,15 +144,17 @@ async def _predict_single(
     if model:
         provider.model = model
     version = prompt_version or "v1"
+    tpl_hash = _prompt_template_hash(version)
 
-    # 0. 查缓存(同 match+provider+model+version 5 分钟内直接返)
-    cached = _get_cached(match_id, settings.LLM_PROVIDER, provider.model, version)
-    if cached is not None:
-        logger.debug("predict cache hit match=%s", match_id)
-        return cached
+    # 0. 查缓存(同 match+provider+model+version+模板hash 5 分钟内直接返)
+    if use_cache:
+        cached = _get_cached(match_id, settings.LLM_PROVIDER, provider.model, version, tpl_hash)
+        if cached is not None:
+            logger.debug("predict cache hit match=%s", match_id)
+            return cached
 
-    # 1. 拼上下文
-    ctx = await build_context(match_id)
+    # 1. 拼上下文(P2-6: backtest 时使用 match_date-1天 作为 cutoff)
+    ctx = await build_context(match_id, backtest=backtest)
 
     # 1.5 计算快照元数据(用于可复现性)
     now = datetime.now(timezone.utc)
@@ -191,6 +232,7 @@ async def _predict_single(
             raw=resp.raw,
         )
 
-        # 5. 写入缓存
-        _set_cached(match_id, settings.LLM_PROVIDER, provider.model, version, result)
+        # 5. 写入缓存(仅当允许缓存时)
+        if use_cache:
+            _set_cached(match_id, settings.LLM_PROVIDER, provider.model, version, tpl_hash, result)
         return result

src/llm/utils.py

@@ -0,0 +1,19 @@
+"""LLM 模块共享工具函数。"""
+from __future__ import annotations
+
+
+def actual_1x2(home: int, away: int) -> str:
+    """实际比分 → 胜平负。
+
+    单一权威源: backtest.py 和 eval.py 共用,避免重复定义。
+    """
+    if home > away:
+        return "1"
+    if home == away:
+        return "X"
+    return "2"
+
+
+def is_correct_1x2(pred: str | None, actual: str) -> bool:
+    """预测是否命中胜平负。"""
+    return pred == actual

src/llm/validation.py

@@ -10,6 +10,11 @@ from pydantic import BaseModel, Field, field_validator, model_validator
 
 logger = logging.getLogger(__name__)
 
+# 单一权威源:从 orchestrator.SPECIALIST_SPECS 派生,避免两端独立定义导致静默偏离
+from src.llm.agents.orchestrator import SPECIALIST_SPECS
+
+KNOWN_AGENT_NAMES: tuple[str, ...] = tuple(spec.name for spec in SPECIALIST_SPECS)
+
 
 class AgentReportSchema(BaseModel):
     """单个专家 Agent 输出的校验 schema。"""
@@ -64,9 +69,17 @@ class PredictionOutputSchema(BaseModel):
 
     @model_validator(mode="after")
     def check_consistency(self) -> "PredictionOutputSchema":
-        """验证比分与胜平负一致,不一致则自动修正。"""
+        """验证比分与胜平负一致。
+
+        不一致时以比分为准修正 pred_1x2(比分是更结构化的输出),
+        但**必须告警** —— 静默修正会掩盖 LLM 的自相矛盾,让问题无法被发现。
+        """
         expected = _score_to_1x2(self.pred_home_goals, self.pred_away_goals)
         if self.pred_1x2 != expected:
+            logger.debug(
+                "1x2 与比分不一致: 比分 %.1f-%.1f 推出 '%s',但 LLM 给出 '%s';以比分修正",
+                self.pred_home_goals, self.pred_away_goals, expected, self.pred_1x2,
+            )
             self.pred_1x2 = expected
         return self
 
@@ -80,6 +93,63 @@ def _score_to_1x2(home: float, away: float) -> str:
     return "X"
 
 
+class AgentWeightsSchema(BaseModel):
+    """终裁给出的各专家权重校验 schema。
+
+    权重含义:各专家报告在最终决策中的相对影响力。约束:
+      - key 必须是已知的 5 个专家名
+      - value ∈ [0, 1]
+      - 总和允许有 0.05 的浮点误差(LLM 常凑不到精确 1.0),
+        超出则归一化到 1.0 而不是直接拒收
+    """
+    weights: dict[str, float] = Field(default_factory=dict)
+
+    @field_validator("weights", mode="before")
+    @classmethod
+    def coerce_weights(cls, v):
+        if v is None:
+            return {}
+        if not isinstance(v, dict):
+            raise ValueError(f"agent_weights 必须是 dict,得到 {type(v).__name__}")
+        out: dict[str, float] = {}
+        for k, raw in v.items():
+            key = str(k).strip().lower()
+            if key not in KNOWN_AGENT_NAMES:
+                logger.warning("agent_weights 含未知专家 '%s',已忽略", k)
+                continue
+            f = _safe_float(raw)
+            if f is None:
+                logger.warning("agent_weights['%s']=%r 非数值,已忽略", k, raw)
+                continue
+            # 负数直接钳到 0;超过 1 的钳到 1
+            out[key] = min(max(f, 0.0), 1.0)
+        return out
+
+    @model_validator(mode="after")
+    def normalize_sum(self) -> "AgentWeightsSchema":
+        """权重和不为 1 时归一化(而非拒收),并在偏离较大时告警。"""
+        if not self.weights:
+            return self
+        total = sum(self.weights.values())
+        if total <= 0:
+            return self
+        if abs(total - 1.0) > 0.05:
+            logger.warning("agent_weights 总和为 %.3f,已归一化到 1.0", total)
+            self.weights = {k: v / total for k, v in self.weights.items()}
+        return self
+
+
+def validate_agent_weights(raw) -> dict[str, float]:
+    """校验并规范化终裁给出的 agent_weights。非法输入返回空 dict。"""
+    if raw is None:
+        return {}
+    try:
+        return AgentWeightsSchema(weights=raw).weights
+    except Exception as e:
+        logger.warning("agent_weights 校验失败,丢弃: %s", e)
+        return {}
+
+
 def validate_agent_output(raw: dict) -> AgentReportSchema:
     """校验并规范化单个 Agent 输出。"""
     return AgentReportSchema(

tests/test_agents.py

@@ -219,3 +219,100 @@ class TestOrchestratorAggregation:
         )
         assert "{{match_header}}" not in rendered
         assert "{{agent_reports}}" not in rendered
+
+
+class TestSliceResultGate:
+    """P2-1: 结构化 has_data 门控(替代脆弱的文案子串匹配)。"""
+
+    def test_sliceresult_empty_is_no_data(self):
+        from src.llm.agents.base import _slice_has_data
+        from src.llm.context_builder import SliceResult
+
+        text, has = _slice_has_data(SliceResult(text="── x ──\n  无数据", has_data=False))
+        assert has is False
+
+    def test_sliceresult_with_data_beats_text(self):
+        """即使文案里出现「无数据」字样,结构化 has_data=True 也应胜出。
+
+        这正是旧实现的漏洞:文案匹配会把「主队: 无伤停数据 / 客队: 2人伤停」
+        这类混合输出……这里显式验证结构化声明优先。
+        """
+        from src.llm.agents.base import _slice_has_data
+        from src.llm.context_builder import SliceResult
+
+        tricky = SliceResult(
+            text="── 阵容完整性 ──\n  主队: 无伤停数据\n  客队伤停(1人):\n    - X: 拉伤",
+            has_data=True,
+        )
+        _, has = _slice_has_data(tricky)
+        assert has is True
+
+    def test_str_fallback_still_works(self):
+        """旧式 str 切片(测试 mock / 自定义切片)仍走文案回退,保持兼容。"""
+        from src.llm.agents.base import _slice_has_data
+
+        assert _slice_has_data("── 伤停 ──\n  无数据")[1] is False
+        assert _slice_has_data("── 交锋 ──\n  A 2-1 B")[1] is True
+
+    def test_sliceresult_str_compat(self):
+        """SliceResult 可当 str 用(老调用点无需改)。"""
+        from src.llm.context_builder import SliceResult
+
+        s = SliceResult(text="hello", has_data=True)
+        assert str(s) == "hello"
+
+
+class TestAgentWeightsValidation:
+    """P2-2: agent_weights 必须过校验才能落库。"""
+
+    def test_unknown_agent_dropped(self):
+        from src.llm.validation import validate_agent_weights
+
+        w = validate_agent_weights({"form": 0.4, "h2h": 0.4, "bogus": 0.2, "stats": 0.4})
+        assert "bogus" not in w
+        assert set(w) <= {"form", "stats", "home_away", "injuries", "h2h"}
+
+    def test_out_of_range_clamped(self):
+        from src.llm.validation import validate_agent_weights
+
+        w = validate_agent_weights({"form": 5.0, "h2h": -1.0})
+        assert w["form"] == 1.0
+        assert w["h2h"] == 0.0
+
+    def test_sum_normalized(self):
+        from src.llm.validation import validate_agent_weights
+
+        w = validate_agent_weights({"form": 2.0, "stats": 2.0})
+        assert abs(sum(w.values()) - 1.0) < 1e-9
+
+    def test_no_weights_returns_empty(self):
+        from src.llm.validation import validate_agent_weights
+
+        assert validate_agent_weights(None) == {}
+        assert validate_agent_weights("not a dict") == {}
+
+
+class TestPredictionConsistencyWarn:
+    """P2-3: 比分与 1x2 不一致 → 以比分修正(且告警)。"""
+
+    def test_mismatch_is_corrected_to_score(self):
+        from src.llm.validation import validate_prediction_output
+
+        v = validate_prediction_output({
+            "pred_home_goals": 2.0,
+            "pred_away_goals": 1.0,
+            "pred_1x2": "X",   # 与 2-1 矛盾
+            "subjective_confidence": 0.7,
+        })
+        assert v.pred_1x2 == "1"   # 按比分修正
+
+    def test_consistent_passes_through(self):
+        from src.llm.validation import validate_prediction_output
+
+        v = validate_prediction_output({
+            "pred_home_goals": 0.0,
+            "pred_away_goals": 0.0,
+            "pred_1x2": "X",
+            "subjective_confidence": 0.5,
+        })
+        assert v.pred_1x2 == "X"

tests/test_regressions.py

@@ -0,0 +1,99 @@
+"""回归测试:锁定 P0 三项「静默失效」缺陷不再复发。
+
+这些用例不依赖数据库 —— 它们用静态分析检查代码结构,
+因为三个 P0 的本质都是「集成方式错误」,在 mock 掉 slice_fn /
+provider 的单元测试里永远照不出来(这正是它们当初漏网的原因)。
+
+- P0-1/P0-2: 查询 Match 的函数必须 eager-load 切片会访问的关系
+- P0-3: bzzoiro 写 source_event_id 时用的 raw 必须与 nm 配对
+"""
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+SRC = Path(__file__).resolve().parent.parent / "src"
+
+# 切片函数会读取的关系属性 → 查询时必须 eager-load
+MATCH_RELATIONS = ("stats", "home_team", "away_team", "league")
+
+
+def _read(rel: str) -> str:
+    return (SRC / rel).read_text(encoding="utf-8")
+
+
+class TestEagerLoadCoverage:
+    """P0-1 / P0-2: 凡是 select(Match) 且后续访问关系的函数,必须有 eager-load。"""
+
+    def test_context_builder_getters_eager_load(self):
+        """_get_form / _get_h2h / _get_home_away 必须预加载关系。
+
+        models.py 已声明 lazy="selectin" 兜底,但这里同时检查显式
+        selectinload —— 显式声明是查询意图的固化,也被 P0 修复所依赖。
+        """
+        src = _read("llm/context_builder.py")
+        for fn in ("_get_form", "_get_h2h", "_get_home_away"):
+            # 截取函数体
+            m = re.search(rf"async def {fn}\(.*?(?=\nasync def |\n# =|\Z)", src, re.S)
+            assert m, f"{fn} 未找到"
+            body = m.group(0)
+            assert "selectinload" in body, (
+                f"{fn} 查询 Match 但未 eager-load 关系 —— "
+                "this would raise MissingGreenlet in async SQLAlchemy (P0-2)"
+            )
+
+    def test_backtest_candidates_eager_load(self):
+        """回测取历史比赛必须 eager-load(否则 session 关闭后访问关系必炸)。"""
+        src = _read("llm/backtest.py")
+        assert "selectinload" in src, "backtest 未 eager-load 关系 (P0-1)"
+
+    def test_relationship_default_is_selectin(self):
+        """models.py 中 Match 的高频关系应声明 lazy='selectin' 作为兜底。"""
+        src = _read("db/models.py")
+        # 找到 Match 类定义段
+        m = re.search(r"class Match\(Base\):.*?(?=\nclass )", src, re.S)
+        assert m, "Match 类未找到"
+        body = m.group(0)
+        for rel in MATCH_RELATIONS:
+            # 关系声明可能跨多行(stats/home_team/away_team 都是),因此按
+            # 「从 `rel: Mapped` 到下一个 `xxx: Mapped` 之前」整段匹配。
+            m_rel = re.search(
+                rf"^\s*{rel}: Mapped.*?(?=^\s*\w+: Mapped|\Z)", body, re.M | re.S
+            )
+            assert m_rel, f"Match.{rel} 未找到"
+            assert 'lazy="selectin"' in m_rel.group(0), (
+                f"Match.{rel} 未声明 lazy='selectin' —— 兜底缺失 (P0-2)"
+            )
+
+
+class TestBzzoiroLineage:
+    """P0-3: source_event_id 必须取配对的 raw,不能是循环残留变量。"""
+
+    def test_normalized_matches_carries_raw(self):
+        src = _read("data/bzzoiro.py")
+        # 规范化结果必须与原始 event 成对保存
+        assert "normalized_matches.append((nm, raw))" in src, (
+            "normalized_matches 未携带 (nm, raw) 元组 —— raw 变量泄漏会回归 (P0-3)"
+        )
+        # 内层消费循环必须解包成对变量
+        assert re.search(r"for nm, raw in normalized_matches", src), (
+            "消费循环未解包 (nm, raw) —— 血缘字段会取到错误 event (P0-3)"
+        )
+
+    def test_no_orphan_raw_use(self):
+        """source_event_id 所在行必须在解包循环内(用缩进 + 上下文粗判)。"""
+        src = _read("data/bzzoiro.py")
+        lines = src.splitlines()
+        # 找到 "for nm, raw in normalized_matches" 所在行号
+        start = next(
+            (i for i, ln in enumerate(lines) if "for nm, raw in normalized_matches" in ln),
+            None,
+        )
+        assert start is not None
+        # 该循环之后、下一个同/更低缩进的顶层语句之前的范围
+        seg = "\n".join(lines[start:])
+        uses = [ln for ln in seg.splitlines() if "source_event_id" in ln]
+        assert uses, "未找到 source_event_id 赋值"
+        assert all("raw.get(" in ln for ln in uses), (
+            "source_event_id 未使用配对的 raw (P0-3)"
+        )