feat: 2026-04-15~05-02 累积变更基线 — AI 重构 + Runtime Context + DWS 修复

涵盖（每条对应已存的审计记录）：
- AI 模块拆分：apps/backend/app/ai/apps -> prompts/（8 个 APP + app2a 派生）
  audit: 2026-04-20__ai-module-complete.md
- admin-web AI 管理套件：AIDashboard / AIOperations / AIRunLogs / AITriggers / TriggerManager
  audit: 2026-04-21__admin-web-ai-management-suite.md
- App2 财务洞察 prompt v3 -> v5.1 + 小程序 AI 接入（chat / board-finance）
  audit: 2026-04-22__app2_prompt_v5_1_and_miniprogram_ai_insight.md
- App2 prewarm 全过滤器 + AI 触发器 cron reschedule
  audit: 2026-04-21__app2-finance-prewarm-all-filters.md
  migration: 20260420_ai_trigger_jobs_and_app2_prewarm.sql / 20260421_app2_prewarm_cron_reschedule.sql
- AppType 联合类型对齐 + adminAiAppTypes.test.ts
  audit: 2026-04-30__admin_web_ai_app_type_alignment.md
- DashScope tokens_used 提取修复
  audit: 2026-04-30__backend_dashscope_tokens_used_extraction.md
- App3 线索完整详情 prompt
  audit: 2026-05-01__backend_app3_full_detail_prompt.md
- Runtime Context 沙箱（5-1~5-2 主线）：
  - 后端 schema/service + admin_runtime_context / xcx_runtime_clock 两个 router
  - admin-web RuntimeContext.tsx + miniprogram runtime-clock.ts
  - migration: 20260501__runtime_context_sandbox.sql
  - tools/db/verify_admin_web_sandbox.py + verify_sandbox_end_to_end.py
  - database/changes: 7 份 sandbox_* 验证报告
- 飞球 DWS 修复：finance_area_daily 区域汇总 + task_engine 调整
  + RLS 视图业务日上界（migration 20260502 + scripts/ops/gen_rls_business_date_migration.py）

合规：
- .gitignore 启用 tmp/ 排除
- 不入仓：apps/etl/connectors/feiqiu/.env（API_TOKEN secret，本地修改保留）

待验证清单：
- docs/audit/changes/2026-05-04__cumulative_baseline_pending_verification.md
  每个主题的功能完整性 / 上线验证几乎都未收口，按优先级 P0~P3 逐一处理

apps/backend/app/routers/xcx_chat.py

@@ -244,27 +244,56 @@ async def chat_stream(
             )
 
             # 流式调用 DashScope Application API
-            async for chunk in client.call_app_stream(
+            # 返回 (text_chunk, session_id_or_none) 元组：累积最后一次 session_id 用于回写
+            latest_session_id: str | None = session_id
+            async for chunk_text, chunk_session_id in client.call_app_stream(
                 app_id=config.app_id_1_chat,
                 prompt=prompt,
                 session_id=session_id,
                 biz_params=biz_params,
             ):
-                full_reply_parts.append(chunk)
+                if chunk_session_id:
+                    latest_session_id = chunk_session_id
+                if not chunk_text:
+                    continue
+                full_reply_parts.append(chunk_text)
                 tokens_total += 1
                 # SSE trace: 每 10 个 token 记录一次
                 record_sse_token(token_count=1, total_tokens=tokens_total)
-                yield f"event: message\ndata: {json.dumps({'token': chunk}, ensure_ascii=False)}\n\n"
+                yield f"event: message\ndata: {json.dumps({'token': chunk_text}, ensure_ascii=False)}\n\n"
 
             # 流结束：拼接完整回复并持久化
             full_reply = "".join(full_reply_parts)
             estimated_tokens = len(full_reply)
 
+            # Phase 1.3：assistant 消息挂 reference_card（若用户从特定详情页入口发起对话）
+            try:
+                from app.ai.references import build_app1_reference_card
+                _ref_card = None
+                _pc = body.page_context or {}
+                _ctx_id = _pc.get("contextId") or _pc.get("taskId") or _pc.get("customerId") or _pc.get("coachId")
+                if body.source_page and _ctx_id:
+                    _ref_card = build_app1_reference_card(body.source_page, _ctx_id)
+            except Exception:
+                logger.warning("构建 reference_card 失败", exc_info=True)
+                _ref_card = None
+
             ai_msg_id, ai_created_at = svc._save_message(
-                body.chat_id, "assistant", full_reply, tokens_used=estimated_tokens,
+                body.chat_id, "assistant", full_reply,
+                tokens_used=estimated_tokens,
+                reference_card=_ref_card,
             )
             svc._update_session_metadata(body.chat_id, full_reply)
 
+            # multi-turn 启用：回写百炼返回的 session_id（若首次对话或服务端更新）
+            if latest_session_id and latest_session_id != session_id:
+                try:
+                    svc.save_session_id(body.chat_id, latest_session_id)
+                except Exception:
+                    logger.warning(
+                        "save_session_id 失败 chat_id=%s", body.chat_id, exc_info=True,
+                    )
+
             # 发送 done 事件
             done_data = json.dumps(
                 {"messageId": ai_msg_id, "createdAt": ai_created_at},