修复：CI 部署环境与 ref 错配、迁移碎片化、图片意图 source_span、章节物化脏版式、会话历史与本地语音不一致

新增：TTS 上传 COS 与分片、章节 reading_segments 物化与快照、markdown 清洗、会话消息 repository、语音 store 重构与相关测试

api/app/features/conversation/models.py

@@ -1,4 +1,13 @@
-from sqlalchemy import Boolean, Column, DateTime, ForeignKey, Integer, String, Text
+from sqlalchemy import (
+    JSON,
+    Boolean,
+    Column,
+    DateTime,
+    ForeignKey,
+    Integer,
+    String,
+    Text,
+)
 from sqlalchemy.orm import relationship
 
 from app.core.db import Base, utc_now
@@ -32,9 +41,11 @@ class Segment(Base):
     conversation_id = Column(String, ForeignKey("conversations.id"), nullable=False)
     audio_url = Column(String, nullable=True)
     transcript_text = Column(Text, nullable=False)
+    audio_duration_seconds = Column(Integer, nullable=True)
     created_at = Column(DateTime(timezone=True), default=utc_now)
     processed = Column(Boolean, default=False)
     topic_category = Column(String, nullable=True)
     agent_response = Column(Text, nullable=True)
+    tts_audio_urls = Column(JSON, nullable=True)
 
     conversation = relationship("Conversation", back_populates="segments")