diff --git a/backend/task_scheduler/app/main.py b/backend/task_scheduler/app/main.py
index 2877534..b8d4dff 100644
--- a/backend/task_scheduler/app/main.py
+++ b/backend/task_scheduler/app/main.py
@@ -466,55 +466,103 @@ async def _async_do_signin(account_id: str, cron_expr: str = ""):
 
 
 async def _fetch_topics(cookies: dict) -> list:
-    """获取关注的超话列表。Cookie 失效时返回空列表。"""
+    """
+    获取关注的超话列表。
+    带重试机制：微博凌晨可能触发风控（302 到通行证验证页），重试 2 次。
+    返回空列表表示获取失败（Cookie 失效或风控）。
+    """
     import httpx
 
-    topics = []
-    try:
-        async with httpx.AsyncClient(timeout=15, follow_redirects=True) as client:
-            resp = await client.get("https://weibo.com/", headers=WEIBO_HEADERS, cookies=cookies)
-            # 检测是否被重定向到访客页（Cookie 失效）
-            if "passport.weibo.com" in str(resp.url) or "visitor" in str(resp.url).lower():
-                logger.warning("Cookie 已失效，被重定向到登录/访客页")
-                return []
+    max_retries = 3
+    for attempt in range(max_retries):
+        topics = []
+        try:
+            # 不自动跟随重定向，手动检测 302
+            async with httpx.AsyncClient(timeout=15, follow_redirects=False) as client:
+                resp = await client.get("https://weibo.com/", headers=WEIBO_HEADERS, cookies=cookies)
 
-            xsrf = client.cookies.get("XSRF-TOKEN", "")
-            headers = {**WEIBO_HEADERS, "X-Requested-With": "XMLHttpRequest"}
-            if xsrf:
-                headers["X-XSRF-TOKEN"] = xsrf
+                # 302 重定向到登录页 = Cookie 失效或风控
+                if resp.status_code in (301, 302):
+                    location = resp.headers.get("location", "")
+                    if "login.sina.com.cn" in location or "passport" in location:
+                        if attempt < max_retries - 1:
+                            wait = (attempt + 1) * 3
+                            logger.warning(f"被重定向到登录页(尝试 {attempt+1}/{max_retries})，{wait}秒后重试...")
+                            await asyncio.sleep(wait)
+                            continue
+                        else:
+                            logger.warning(f"多次重试仍被重定向，Cookie 可能已失效")
+                            return []
+
+                    # 其他 302（如 weibo.com → www.weibo.com），跟随一次
+                    resp = await client.get(location, headers=WEIBO_HEADERS, cookies=cookies)
+
+                xsrf = ""
+                for c in resp.cookies.jar:
+                    if c.name == "XSRF-TOKEN":
+                        xsrf = c.value
+                        break
+
+                headers = {**WEIBO_HEADERS, "X-Requested-With": "XMLHttpRequest"}
+                if xsrf:
+                    headers["X-XSRF-TOKEN"] = xsrf
+
+                page = 1
+                while page <= 10:
+                    resp = await client.get(
+                        "https://weibo.com/ajax/profile/topicContent",
+                        params={"tabid": "231093_-_chaohua", "page": str(page)},
+                        headers=headers, cookies=cookies,
+                    )
+
+                    # 超话接口也可能被 302
+                    if resp.status_code in (301, 302):
+                        if attempt < max_retries - 1:
+                            wait = (attempt + 1) * 3
+                            logger.warning(f"超话接口被重定向(尝试 {attempt+1}/{max_retries})，{wait}秒后重试...")
+                            await asyncio.sleep(wait)
+                            break  # break inner loop, retry outer
+                        else:
+                            logger.warning("超话接口多次重定向，放弃")
+                            return []
+
+                    try:
+                        data = resp.json()
+                    except Exception:
+                        logger.warning(f"超话列表响应非 JSON: {resp.text[:200]}")
+                        break
+                    if data.get("ok") != 1:
+                        break
+                    tlist = data.get("data", {}).get("list", [])
+                    if not tlist:
+                        break
+                    for item in tlist:
+                        title = item.get("topic_name", "") or item.get("title", "")
+                        cid = ""
+                        for field in ("oid", "scheme"):
+                            m = re.search(r"100808[0-9a-fA-F]+", item.get(field, ""))
+                            if m:
+                                cid = m.group(0)
+                                break
+                        if title and cid:
+                            topics.append({"title": title, "containerid": cid})
+                    if page >= data.get("data", {}).get("max_page", 1):
+                        break
+                    page += 1
+                else:
+                    # while 正常结束（没有 break），说明成功
+                    if topics:
+                        return topics
+
+                # 如果有 topics 说明成功了
+                if topics:
+                    return topics
+
+        except Exception as e:
+            logger.error(f"获取超话列表失败(尝试 {attempt+1}/{max_retries}): {e}")
+            if attempt < max_retries - 1:
+                await asyncio.sleep(3)
 
-            page = 1
-            while page <= 10:
-                resp = await client.get(
-                    "https://weibo.com/ajax/profile/topicContent",
-                    params={"tabid": "231093_-_chaohua", "page": str(page)},
-                    headers=headers, cookies=cookies,
-                )
-                try:
-                    data = resp.json()
-                except Exception:
-                    logger.warning(f"超话列表响应非 JSON: {resp.text[:200]}")
-                    break
-                if data.get("ok") != 1:
-                    break
-                tlist = data.get("data", {}).get("list", [])
-                if not tlist:
-                    break
-                for item in tlist:
-                    title = item.get("topic_name", "") or item.get("title", "")
-                    cid = ""
-                    for field in ("oid", "scheme"):
-                        m = re.search(r"100808[0-9a-fA-F]+", item.get(field, ""))
-                        if m:
-                            cid = m.group(0)
-                            break
-                    if title and cid:
-                        topics.append({"title": title, "containerid": cid})
-                if page >= data.get("data", {}).get("max_page", 1):
-                    break
-                page += 1
-    except Exception as e:
-        logger.error(f"获取超话列表失败: {e}")
     return topics
 
 
diff --git a/frontend/Dockerfile b/frontend/Dockerfile
index e8ab7a3..08ddc60 100644
--- a/frontend/Dockerfile
+++ b/frontend/Dockerfile
@@ -32,6 +32,6 @@ USER appuser
 EXPOSE 5000
 
 HEALTHCHECK --interval=30s --timeout=10s --retries=3 \
-    CMD curl -f http://localhost:5000/ || exit 1
+    CMD curl -f http://localhost:5000/health || exit 1
 
 CMD ["python", "app.py"]
diff --git a/frontend/app.py b/frontend/app.py
index 32e0fac..4f6dac3 100644
--- a/frontend/app.py
+++ b/frontend/app.py
@@ -105,6 +105,10 @@ def index():
         return redirect(url_for('dashboard'))
     return redirect(url_for('login'))
 
+@app.route('/health')
+def health():
+    return 'ok', 200
+
 @app.route('/register', methods=['GET', 'POST'])
 def register():
     if request.method == 'POST':