fix(capabilities): redetect 同步刷新 app.state 快照并升级 tickflow SDK 0.1.25

- /api/capabilities/redetect 重新探测后同步写回 app.state.capabilities,
  minute_refresh 等门控服务即时读到新档位, 不再需要重启
- tickflow 0.1.24 缺 intraday_universe 方法导致 Expert 档全量分钟探测失败,
  升级到 0.1.25
- 全量分钟轮次异常与模式补充 warning 日志, 便于线上诊断

backend/app/api/routes.py

@@ -1,7 +1,7 @@
 """API 路由 — Phase 0 仅 /health 与 /api/capabilities。"""
 from __future__ import annotations
 
-from fastapi import APIRouter
+from fastapi import APIRouter, Request
 
 from app import __version__
 from app.tickflow import client as tf_client
@@ -31,9 +31,14 @@ def capabilities() -> dict:
 
 
 @router.post("/api/capabilities/redetect")
-def redetect() -> dict:
+def redetect(request: Request) -> dict:
     """用户在设置页"重新检测"按钮。"""
     capset = detect_capabilities(force=True)
+    # 同步刷新 app.state 快照 (minute_refresh 等服务的门控读这里) 与财务调度器,
+    # 与 settings.py 各探测路径一致 — 否则重检测后服务侧仍读旧 capset 被错误门控
+    request.app.state.capabilities = capset
+    from app.api.settings import _sync_financial_scheduler_caps
+    _sync_financial_scheduler_caps(request.app.state, capset)
     return {
         "label": tier_label(),
         "capabilities": capset.to_dict(),

backend/app/services/minute_refresh.py

@@ -34,6 +34,7 @@
 from __future__ import annotations
 
 import contextlib
+import logging
 import threading
 import time
 from dataclasses import dataclass, field
@@ -44,6 +45,8 @@ import polars as pl
 from app.market_time import cn_now, cn_today, in_continuous_session
 from app.services import preferences
 
+logger = logging.getLogger(__name__)
+
 # 轮询间隔允许范围 (秒): 稳态轮单请求无并发脉冲, 下限 3s;
 # 上限 120s — universe 端点每标的只回最新 3 根, 间隔超过 3 分钟必留缺口,
 # 每轮都会触发修复轮, 稳态设计失效, 故不允许配到 120s 以上。
@@ -187,6 +190,7 @@ class MinuteRefreshService:
                     continue
             except Exception as e:
                 self._state.last_error = f"round failed: {e}"
+                logger.warning("全量分钟轮次异常: %s", e)
             self._stop.wait(_LOOP_STEP_S)
 
     # ------------------------------------------------------------------
@@ -231,8 +235,11 @@ class MinuteRefreshService:
 
         t0 = time.perf_counter()
         mode = self._select_mode()
+        mode_label = "增量" if mode == "increment" else "全天修复"
         with self._round_lock:
             if mode == "increment":
+                # fetch 计时只覆盖网络取数; full 分支的 universe 维表读取不计入
+                fetch_started = time.perf_counter()
                 df, requests = kline_sync.fetch_intraday_universe_increment()
                 self._state.last_symbols = (
                     df["symbol"].n_unique() if not df.is_empty() else 0
@@ -242,18 +249,27 @@ class MinuteRefreshService:
                 self._state.last_symbols = len(symbols)
                 if not symbols:
                     self._state.last_error = "empty universe (instruments 未加载)"
+                    logger.warning("全量分钟[%s] 本轮中止: 标的池为空 (instruments 未加载)", mode_label)
                     return
                 capset = getattr(self._app_state, "capabilities", None) if self._app_state else None
+                fetch_started = time.perf_counter()
                 df, requests = kline_sync.fetch_intraday_full_market_burst(symbols, capset)
+            fetch_ms = (time.perf_counter() - fetch_started) * 1000
             self._state.last_requests = requests
             if df.is_empty():
                 self._empty_rounds += 1
                 self._state.last_error = f"intraday {mode} returned no data"
+                logger.warning(
+                    "全量分钟[%s] 本轮返回空数据: %d 请求, 取数 %.0fms",
+                    mode_label, requests, fetch_ms,
+                )
                 return
             self._empty_rounds = 0
+            write_started = time.perf_counter()
             written = kline_sync._write_minute_partition(
                 df, self._repo.store.data_dir / "kline_minute",
             )
+            write_ms = (time.perf_counter() - write_started) * 1000
 
         self._state.rounds += 1
         self._state.last_round_at = time.time()
@@ -261,6 +277,12 @@ class MinuteRefreshService:
         self._state.last_rows = written
         self._state.last_mode = mode
         self._state.last_error = None
+        logger.info(
+            "全量分钟[%s] 第 %d 轮: 取数 %.0fms (%d 请求, %d 标的), "
+            "落盘 %.0fms (%d 行), 总计 %.0fms",
+            mode_label, self._state.rounds, fetch_ms, self._state.last_requests,
+            self._state.last_symbols, write_ms, written, self._state.last_round_ms,
+        )
 
     def _universe(self) -> list[str]:
         """全市场 A 股标的 (instruments 维表, 与盘后分钟同步同一来源)。"""