fix: 修复 SQLite 连接复用导致数据库不可用

# Conflicts:
#	tests/integration/test_routes.py

.gitignore

@@ -8,3 +8,4 @@ htmlcov/
 .env
 data/*.db
 data/*.db-*
+data/corrupt-backups/

Dockerfile

@@ -7,9 +7,8 @@ WORKDIR /app
 
 RUN adduser --disabled-password --gecos "" appuser
 
-COPY pyproject.toml uv.lock ./
-RUN pip install --no-cache-dir uv \
-    && uv pip install --system --no-cache .
+COPY requirements.txt ./
+RUN pip install --no-cache-dir -r requirements.txt
 
 COPY app ./app
 COPY .env.example ./.env.example

app/db.py

@@ -5,6 +5,7 @@ from datetime import UTC, datetime
 
 from sqlalchemy import Engine, create_engine, event
 from sqlalchemy.orm import DeclarativeBase, Session, sessionmaker
+from sqlalchemy.pool import NullPool
 
 from app.config import get_settings
 
@@ -20,7 +21,10 @@ def create_sqlite_engine(database_url: str):
             db_path.parent.mkdir(parents=True, exist_ok=True)
 
     connect_args = {"check_same_thread": False, "timeout": 10}
-    engine = create_engine(database_url, connect_args=connect_args)
+    engine_kwargs = {"connect_args": connect_args}
+    if database_url.startswith("sqlite:///") and database_url != "sqlite:///:memory:":
+        engine_kwargs["poolclass"] = NullPool
+    engine = create_engine(database_url, **engine_kwargs)
     engine.dialect.connect_args = connect_args
 
     @event.listens_for(engine, "connect")

app/main.py

@@ -1,11 +1,9 @@
 from collections.abc import AsyncGenerator
 from contextlib import asynccontextmanager
 import logging
-from urllib.parse import quote
 
 from fastapi import Depends, FastAPI, HTTPException, Request, status
 from fastapi.responses import HTMLResponse, JSONResponse
-from fastapi.responses import Response
 from fastapi.staticfiles import StaticFiles
 from sqlalchemy.exc import OperationalError, SQLAlchemyError
 from sqlalchemy import select
@@ -14,7 +12,6 @@ from sqlalchemy.orm import Session
 from app.db import SessionLocal, backup_sqlite_files, get_db_session, init_db
 from app.models import Comment, ContentItem, Hotspot, Report
 from app.schemas import CreateTaskRequest, CreateTaskResponse, TaskResponse
-from app.services.export_service import export_hotspot_comments_csv, export_item_comments_csv, export_report_markdown
 from app.services.task_service import (
     RUNNING_TASK_MESSAGE,
     create_task,
@@ -22,6 +19,7 @@ from app.services.task_service import (
     has_running_task,
     list_tasks,
     recover_running_tasks,
+    recover_stale_running_tasks,
 )
 from app.templating import templates
 
@@ -86,6 +84,7 @@ def create_task_api(
     request: CreateTaskRequest,
     session: Session = Depends(get_db_session),
 ) -> CreateTaskResponse:
+    recover_stale_running_tasks(session)
     if has_running_task(session):
         raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail=RUNNING_TASK_MESSAGE)
 
@@ -146,47 +145,3 @@ def item_detail_page(item_id: str, request: Request, session: Session = Depends(
         "items/detail.html",
         {"task": task, "hotspot": hotspot, "item": item, "report": report, "comments": comments},
     )
-
-
-@app.get("/api/export/items/{item_id}/comments.csv")
-def export_item_comments(item_id: str, session: Session = Depends(get_db_session)) -> Response:
-    try:
-        content, filename = export_item_comments_csv(session, item_id)
-    except ValueError:
-        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="导出数据不存在")
-    return Response(
-        content=content,
-        media_type="text/csv; charset=utf-8",
-        headers={"Content-Disposition": f"attachment; filename*=UTF-8''{quote(filename)}"},
-    )
-
-
-@app.get("/api/export/hotspots/{hotspot_id}/comments.csv")
-def export_hotspot_comments(hotspot_id: str, session: Session = Depends(get_db_session)) -> Response:
-    try:
-        content, filename = export_hotspot_comments_csv(session, hotspot_id)
-    except ValueError:
-        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="导出数据不存在")
-    return Response(
-        content=content,
-        media_type="text/csv; charset=utf-8",
-        headers={"Content-Disposition": f"attachment; filename*=UTF-8''{quote(filename)}"},
-    )
-
-
-@app.get("/api/export/hotspots/{hotspot_id}.md")
-def export_hotspot_markdown(hotspot_id: str, session: Session = Depends(get_db_session)) -> Response:
-    try:
-        content, filename = export_report_markdown(session, report_type="hotspot", hotspot_id=hotspot_id)
-    except ValueError:
-        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="报告不存在")
-    return Response(content=content, media_type="text/markdown; charset=utf-8", headers={"Content-Disposition": f"attachment; filename*=UTF-8''{quote(filename)}"})
-
-
-@app.get("/api/export/items/{item_id}.md")
-def export_item_markdown(item_id: str, session: Session = Depends(get_db_session)) -> Response:
-    try:
-        content, filename = export_report_markdown(session, report_type="item", item_id=item_id)
-    except ValueError:
-        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="报告不存在")
-    return Response(content=content, media_type="text/markdown; charset=utf-8", headers={"Content-Disposition": f"attachment; filename*=UTF-8''{quote(filename)}"})

app/platforms/base.py

@@ -87,6 +87,12 @@ class TikHubClient:
                     )
                 time.sleep(2**attempt)
                 continue
+            if response.status_code == 401:
+                raise PlatformAPIError(
+                    "TikHub 鉴权失败，请检查 TIKHUB_API_KEY 是否有效",
+                    error_type="auth_error",
+                    status_code=response.status_code,
+                )
             if response.is_error:
                 raise PlatformAPIError(
                     f"External API returned HTTP {response.status_code}",

app/schemas.py

@@ -34,6 +34,11 @@ class TaskResponse(BaseModel):
     current_stage: str | None = None
     current_stage_label: str | None = None
     last_progress_at: datetime | None = None
+    display_number: int = 0
+    display_id: str = ""
+    created_at_label: str = ""
+    scale_label: str = ""
+    error_summary: str | None = None
     running_seconds: int = 0
     seconds_since_last_progress: int | None = None
     running_duration_label: str = "刚刚"

app/services/ai_service.py

@@ -1,5 +1,6 @@
 import json
 import time
+from concurrent.futures import ThreadPoolExecutor, TimeoutError
 from dataclasses import dataclass
 from typing import Literal
 
@@ -121,12 +122,14 @@ def analyze_comments_with_retry(
     *,
     requester,
     max_retries: int = 3,
+    timeout_seconds: float | None = None,
 ) -> list[AIAnalysisResult]:
     expected_ids = {str(comment["comment_id"]) for comment in comments}
     prompt = build_comment_prompt(comments)
     for attempt in range(max_retries):
         try:
-            return parse_ai_comment_response(requester(prompt), expected_comment_ids=expected_ids)
+            response_text = _request_with_timeout(requester, prompt, timeout_seconds=timeout_seconds)
+            return parse_ai_comment_response(response_text, expected_comment_ids=expected_ids)
         except Exception:
             if attempt >= max_retries - 1:
                 break
@@ -143,6 +146,21 @@ def analyze_comments_with_retry(
     ]
 
 
+def _request_with_timeout(requester, prompt: str, *, timeout_seconds: float | None) -> str:
+    if timeout_seconds is None or timeout_seconds <= 0:
+        return requester(prompt)
+
+    executor = ThreadPoolExecutor(max_workers=1)
+    future = executor.submit(requester, prompt)
+    try:
+        return future.result(timeout=timeout_seconds)
+    except TimeoutError as exc:
+        future.cancel()
+        raise TimeoutError("AI request timed out") from exc
+    finally:
+        executor.shutdown(wait=False, cancel_futures=True)
+
+
 class OpenAICompatibleAIClient:
     def __init__(
         self,

app/services/export_service.py

@@ -1,104 +0,0 @@
-import csv
-import io
-import json
-import re
-
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.models import Comment, ContentItem, Hotspot, Report
-
-
-def labels_to_text(labels_json: str | None) -> str:
-    try:
-        labels = json.loads(labels_json or "[]")
-    except json.JSONDecodeError:
-        return ""
-    return "，".join(str(label) for label in labels) if isinstance(labels, list) else ""
-
-
-def safe_csv_field(value) -> str:
-    text = "" if value is None else str(value)
-    text = text.replace("\r\n", " ").replace("\n", " ").replace("\r", " ")
-    if text.startswith(("=", "+", "-", "@")):
-        return f"'{text}"
-    return text
-
-
-def safe_filename(platform: str, task_id: str, hotspot_keyword: str, *, extension: str = "csv") -> str:
-    keyword = hotspot_keyword[:20] + ("..." if len(hotspot_keyword) > 20 else "")
-    name = f"{platform}_{task_id}_{keyword}"
-    name = re.sub(r'[/\\:*?"<>|]+', "_", name)
-    name = re.sub(r"_+", "_", name).strip("_")
-    return f"{name}.{extension}"
-
-
-def export_item_comments_csv(session: Session, item_id: str) -> tuple[bytes, str]:
-    item = session.get(ContentItem, item_id)
-    if item is None:
-        raise ValueError("content item not found")
-    hotspot = session.get(Hotspot, item.hotspot_id)
-    comments = list(session.scalars(select(Comment).where(Comment.content_item_id == item.id)))
-    output = io.StringIO()
-    writer = csv.writer(output)
-    writer.writerow(["平台", "任务 ID", "热点 ID", "热点标题", "内容条目 ID", "内容条目标题", "评论 ID", "评论内容", "情绪倾向", "方向标签", "点赞数", "评论时间"])
-    for comment in comments:
-        writer.writerow(
-            [
-                item.platform,
-                item.task_id,
-                hotspot.id if hotspot else "",
-                hotspot.title if hotspot else "",
-                item.id,
-                item.title or "",
-                comment.source_comment_id or "",
-                safe_csv_field(comment.content),
-                comment.sentiment,
-                labels_to_text(comment.labels),
-                comment.like_count or 0,
-                comment.comment_time or "",
-            ]
-        )
-    filename = safe_filename(item.platform, item.task_id, hotspot.title if hotspot else item.title or "comments")
-    return output.getvalue().encode("utf-8-sig"), filename
-
-
-def export_hotspot_comments_csv(session: Session, hotspot_id: str) -> tuple[bytes, str]:
-    hotspot = session.get(Hotspot, hotspot_id)
-    if hotspot is None:
-        raise ValueError("hotspot not found")
-    comments = list(session.scalars(select(Comment).where(Comment.hotspot_id == hotspot.id)))
-    output = io.StringIO()
-    writer = csv.writer(output)
-    writer.writerow(["平台", "任务 ID", "热点 ID", "热点标题", "内容条目 ID", "内容条目标题", "评论 ID", "评论内容", "情绪倾向", "方向标签", "点赞数", "评论时间"])
-    for comment in comments:
-        item = session.get(ContentItem, comment.content_item_id)
-        writer.writerow(
-            [
-                comment.platform,
-                comment.task_id,
-                hotspot.id,
-                hotspot.title,
-                item.id if item else "",
-                item.title if item else "",
-                comment.source_comment_id or "",
-                safe_csv_field(comment.content),
-                comment.sentiment,
-                labels_to_text(comment.labels),
-                comment.like_count or 0,
-                comment.comment_time or "",
-            ]
-        )
-    return output.getvalue().encode("utf-8-sig"), safe_filename(hotspot.platform, hotspot.task_id, hotspot.title)
-
-
-def export_report_markdown(session: Session, *, report_type: str, hotspot_id: str | None = None, item_id: str | None = None) -> tuple[str, str]:
-    query = select(Report).where(Report.report_type == report_type)
-    if hotspot_id:
-        query = query.where(Report.hotspot_id == hotspot_id)
-    if item_id:
-        query = query.where(Report.content_item_id == item_id)
-    report = session.scalar(query)
-    if report is None:
-        raise ValueError("report not found")
-    return report.markdown_content or report.markdown, safe_filename("report", report.task_id, report.title, extension="md")

app/services/task_service.py

@@ -3,6 +3,7 @@ import json
 from concurrent.futures import ThreadPoolExecutor
 from collections.abc import Callable
 from datetime import UTC, datetime
+from zoneinfo import ZoneInfo
 
 from sqlalchemy import func, select
 from sqlalchemy.orm import Session, sessionmaker
@@ -21,6 +22,8 @@ from app.services.report_service import generate_hotspot_report, generate_item_r
 RUNNING_TASK_MESSAGE = "当前有正在运行的任务，请稍后再试"
 RESTART_ERROR_MESSAGE = "系统重启，任务被中断"
 STALE_PROGRESS_THRESHOLD_SECONDS = 10 * 60
+STALE_PROGRESS_ERROR_TYPE = "stale_progress_timeout"
+DISPLAY_TIMEZONE = ZoneInfo("Asia/Shanghai")
 task_executor = ThreadPoolExecutor(max_workers=1)
 
 
@@ -34,12 +37,40 @@ STAGE_LABELS = {
     "success": "已完成",
     "failed": "失败",
 }
+PLATFORM_LABELS = {
+    "xiaohongshu": "小红书",
+    "douyin": "抖音",
+}
 
 
 def has_running_task(session: Session) -> bool:
     return session.scalar(select(Task.id).where(Task.status == "running").limit(1)) is not None
 
 
+def recover_stale_running_tasks(session: Session) -> int:
+    now = ensure_utc_datetime(utc_now())
+    tasks = list(session.scalars(select(Task).where(Task.status == "running")))
+    recovered = 0
+    for task in tasks:
+        last_progress_at = ensure_utc_datetime(task.last_progress_at or task.started_at or task.created_at)
+        if last_progress_at is None or now is None:
+            continue
+        seconds_since_progress = max(0, int((now - last_progress_at).total_seconds()))
+        if seconds_since_progress < STALE_PROGRESS_THRESHOLD_SECONDS:
+            continue
+        _mark_task_failed(
+            task,
+            "system",
+            STALE_PROGRESS_ERROR_TYPE,
+            f"任务超过 {STALE_PROGRESS_THRESHOLD_SECONDS // 60} 分钟没有进度更新，已自动标记失败",
+        )
+        task.finished_at = task.finished_at or utc_now()
+        recovered += 1
+    if recovered:
+        session.commit()
+    return recovered
+
+
 def recover_running_tasks(session: Session) -> int:
     tasks = list(session.scalars(select(Task).where(Task.status == "running")))
     for task in tasks:
@@ -150,6 +181,12 @@ def get_task(session: Session, task_id: str) -> Task | None:
 
 
 def hydrate_task_progress(session: Session, task: Task) -> Task:
+    display_number = calculate_task_display_number(session, task)
+    task.display_number = display_number
+    task.display_id = str(display_number)
+    task.created_at_label = format_datetime_minute(task.created_at)
+    task.scale_label = f"{task.hotspot_limit}热点 × {task.item_limit_per_hotspot}内容 × {task.comment_limit_per_item}评论"
+    task.error_summary = build_task_error_summary(task)
     task.hotspots_count = session.scalar(select(func.count(Hotspot.id)).where(Hotspot.task_id == task.id)) or 0
     task.comments_count = session.scalar(select(func.count(Comment.id)).where(Comment.task_id == task.id)) or 0
     task.reports_count = session.scalar(select(func.count(Report.id)).where(Report.task_id == task.id)) or 0
@@ -182,6 +219,29 @@ def hydrate_task_progress(session: Session, task: Task) -> Task:
     return task
 
 
+def calculate_task_display_number(session: Session, task: Task) -> int:
+    task_ids = list(session.scalars(select(Task.id).order_by(Task.created_at, Task.id)))
+    try:
+        return task_ids.index(task.id) + 1
+    except ValueError:
+        return len(task_ids) + 1
+
+
+def format_datetime_minute(value: datetime | None) -> str:
+    value = ensure_utc_datetime(value)
+    return value.astimezone(DISPLAY_TIMEZONE).strftime("%Y-%m-%d %H:%M") if value else ""
+
+
+def build_task_error_summary(task: Task) -> str | None:
+    if task.error_message:
+        return task.error_message
+    if task.error_stage == "system" and task.error_type == "unexpected_restart":
+        return RESTART_ERROR_MESSAGE
+    if task.error_stage or task.error_type:
+        return "任务失败，请查看后端日志"
+    return None
+
+
 def format_duration_zh(total_seconds: int | None) -> str:
     if total_seconds is None:
         return "暂无记录"
@@ -220,6 +280,8 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
         task = session.get(Task, task_id)
         if task is None:
             return
+        task.started_at = task.started_at or utc_now()
+        session.commit()
         try:
             platform = build_platform(task.platform)
             ai_requester, report_summary_provider = build_ai_dependencies()
@@ -309,6 +371,7 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
                             [{"comment_id": comment.id, "content": comment.content} for comment in item_comments],
                             requester=ai_requester,
                             max_retries=get_settings().ai_max_retries,
+                            timeout_seconds=get_settings().ai_timeout_seconds,
                         )
                         results_by_comment_id = {result.comment_id: result for result in ai_results}
                         for comment in item_comments:
@@ -354,9 +417,11 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
                 update_task_stage(task, "success")
             else:
                 _mark_task_failed(task, task.error_stage or "crawl_items", task.error_type or "no_successful_items", task.error_message or "没有任何内容条目成功")
+            task.finished_at = task.finished_at or utc_now()
             session.commit()
         except Exception as exc:
             _mark_task_failed(task, "system", "unexpected_error", str(exc))
+            task.finished_at = task.finished_at or utc_now()
             session.commit()
     finally:
         if close_session:

app/static/app.css

@@ -1,16 +1,63 @@
 body {
-  background: #f5f7fb;
-  color: #172033;
+  background:
+    linear-gradient(180deg, #f4f7fb 0%, #eef3f8 42%, #f8fafc 100%);
+  color: #182235;
+  min-height: 100vh;
 }
 
+.app-navbar {
+  background: rgba(255, 255, 255, 0.94) !important;
+  border-bottom: 1px solid #dfe7f1;
+  box-shadow: 0 10px 30px rgba(19, 34, 56, 0.04);
+}
+
+.navbar-brand {
+  color: #102033;
+}
+
+.nav-link {
+  color: #526173;
+}
+
+.tool-card,
 .card {
+  background: rgba(255, 255, 255, 0.96);
   border-radius: 8px;
-  border: 1px solid #dfe5ef;
-  box-shadow: 0 10px 28px rgba(31, 42, 68, 0.06);
+  border: 1px solid #dbe4ef;
+  box-shadow: 0 16px 42px rgba(31, 42, 68, 0.07);
+  overflow: hidden;
 }
 
-.navbar {
-  background: #fff !important;
+.card-header,
+.section-heading {
+  background: linear-gradient(180deg, #ffffff, #f8fafc);
+  border-bottom: 1px solid #e4ebf3;
+  padding: 18px 20px;
+}
+
+.section-heading {
+  align-items: center;
+  display: flex;
+  justify-content: space-between;
+  gap: 16px;
+}
+
+.section-heading h2,
+.page-heading h1 {
+  font-size: 22px;
+  font-weight: 700;
+  letter-spacing: 0;
+  margin: 0;
+}
+
+.section-kicker,
+.eyebrow {
+  color: #0e766e;
+  font-size: 12px;
+  font-weight: 700;
+  letter-spacing: 0;
+  margin: 0 0 4px;
+  text-transform: uppercase;
 }
 
 .hero-band {
@@ -20,7 +67,9 @@ body {
     url("https://images.unsplash.com/photo-1551288049-bebda4e38f71?auto=format&fit=crop&w=1600&q=80");
   background-position: center;
   background-size: cover;
+  border: 1px solid rgba(255, 255, 255, 0.18);
   border-radius: 8px;
+  box-shadow: 0 24px 60px rgba(15, 36, 62, 0.18);
   color: #fff;
   display: flex;
   justify-content: space-between;
@@ -52,20 +101,67 @@ body {
   gap: 12px;
 }
 
-.eyebrow {
-  color: #71d4c7;
-  font-size: 13px;
-  font-weight: 700;
-  letter-spacing: 0;
-  text-transform: uppercase;
+.page-heading {
+  align-items: center;
+  display: flex;
+  justify-content: space-between;
+  gap: 16px;
 }
 
 .metric-box {
-  background: #f8fafc;
+  background: linear-gradient(180deg, #f9fbfd, #f3f7fb);
   border: 1px solid #e3e9f2;
   border-radius: 8px;
   height: 100%;
-  padding: 14px;
+  padding: 16px;
+}
+
+.metric-box strong {
+  color: #132033;
+  display: block;
+  font-size: 18px;
+  margin: 4px 0;
+}
+
+.app-table thead th {
+  background: #f8fafc;
+  color: #536174;
+  font-size: 12px;
+  font-weight: 700;
+  text-transform: uppercase;
+  white-space: nowrap;
+}
+
+.app-table tbody tr {
+  border-color: #e8eef5;
+}
+
+.app-table tbody tr:hover {
+  background: #f7fbff;
+}
+
+.task-number {
+  align-items: center;
+  background: #e7f0ff;
+  border: 1px solid #cfe0ff;
+  border-radius: 999px;
+  color: #0b5ed7;
+  display: inline-flex;
+  font-weight: 700;
+  height: 32px;
+  justify-content: center;
+  min-width: 32px;
+  padding: 0 10px;
+}
+
+.task-overview {
+  border-top: 3px solid #0d6efd;
+}
+
+.hotspot-item {
+  border-color: #e3eaf2;
+  margin-bottom: 10px;
+  overflow: hidden;
 }
 
 .status-panel {
@@ -98,4 +194,10 @@ body {
   .hero-copy h1 {
     font-size: 36px;
   }
+
+  .section-heading,
+  .page-heading {
+    align-items: flex-start;
+    flex-direction: column;
+  }
 }

app/static/app.js

@@ -111,46 +111,3 @@ function pollTaskListStatus() {
   };
   window.setTimeout(poll, intervalMs);
 }
-
-async function downloadExport(url, defaultFilename, event) {
-  if (event) event.preventDefault();
-  const button = event ? event.currentTarget : null;
-  const originalText = button ? button.textContent : "";
-  if (button) {
-    button.disabled = true;
-    button.textContent = "下载中...";
-  }
-  try {
-    const resp = await fetch(url, { cache: "no-store" });
-    if (!resp.ok) {
-      let message = "导出失败，请稍后重试。";
-      try {
-        const data = await resp.json();
-        if (data.detail) message = data.detail;
-      } catch (_error) {
-        message = resp.status === 503 ? "数据库暂时不可用，请稍后重试。" : message;
-      }
-      alert(message);
-      return;
-    }
-    const blob = await resp.blob();
-    const disposition = resp.headers.get("Content-Disposition") || "";
-    const match = disposition.match(/filename\*=UTF-8''([^;]+)/);
-    const filename = match ? decodeURIComponent(match[1]) : defaultFilename;
-    const blobUrl = URL.createObjectURL(blob);
-    const link = document.createElement("a");
-    link.href = blobUrl;
-    link.download = filename;
-    document.body.appendChild(link);
-    link.click();
-    document.body.removeChild(link);
-    URL.revokeObjectURL(blobUrl);
-  } catch (_error) {
-    alert("网络异常，导出未完成。");
-  } finally {
-    if (button) {
-      button.disabled = false;
-      button.textContent = originalText;
-    }
-  }
-}

app/templates/base.html

@@ -10,9 +10,9 @@
   <script src="/static/app.js" defer></script>
 </head>
 <body>
-  <nav class="navbar navbar-expand-lg bg-body-tertiary border-bottom">
+  <nav class="navbar navbar-expand-lg app-navbar">
     <div class="container">
-      <a class="navbar-brand" href="/">热榜评论分析工具</a>
+      <a class="navbar-brand fw-semibold" href="/">热榜评论雷达</a>
       <div class="navbar-nav">
         <a class="nav-link" href="/">任务列表</a>
       </div>
@@ -28,8 +28,5 @@
     {% endblock %}
     {% block content %}{% endblock %}
   </main>
-  <footer class="border-top py-3">
-    <div class="container text-muted small">内部演示工具 | 仅供学习参考</div>
-  </footer>
 </body>
 </html>

app/templates/hotspots/report.html

@@ -3,26 +3,13 @@
 {% block breadcrumbs %}
 <nav aria-label="breadcrumb"><ol class="breadcrumb">
   <li class="breadcrumb-item"><a href="/">首页</a></li>
-  <li class="breadcrumb-item"><a href="/tasks/{{ task.id }}">任务 #{{ task.id }}</a></li>
+  <li class="breadcrumb-item"><a href="/tasks/{{ task.id }}">任务 {{ task.display_id }}</a></li>
   <li class="breadcrumb-item active">汇总报告</li>
 </ol></nav>
 {% endblock %}
 {% block content %}
 <div class="d-flex justify-content-between align-items-center mb-3">
   <h1 class="h3 mb-0">{{ hotspot.title }} 汇总报告</h1>
-  <div class="d-flex gap-2">
-    <button
-      class="btn btn-outline-primary"
-      onclick="downloadExport('/api/export/hotspots/{{ hotspot.id }}.md', 'hotspot-report.md', event)"
-      {% if not report %}disabled title="报告尚未生成"{% endif %}>
-      导出 Markdown 报告
-    </button>
-    <button
-      class="btn btn-outline-secondary"
-      onclick="downloadExport('/api/export/hotspots/{{ hotspot.id }}/comments.csv', 'hotspot-comments.csv', event)">
-      导出热点评论 CSV
-    </button>
-  </div>
 </div>
 {% if report %}
   {% include "partials/report_panel.html" %}
@@ -30,7 +17,6 @@
   <div class="text-center text-muted py-5">
     <div class="spinner-border text-warning mb-3" role="status"></div>
     <p>报告生成中，请稍候...</p>
-    <p class="small mb-0">Markdown 报告将在分析完成后开放下载；评论 CSV 可先导出已抓取的数据。</p>
   </div>
 {% endif %}
 {% endblock %}

app/templates/index.html

@@ -8,14 +8,16 @@
     <h1>热榜评论雷达</h1>
     <p>从小红书和抖音热点出发，抓取真实评论，生成 AI 情绪、标签和可导出的分析报告。</p>
   </div>
-  <div class="hero-actions">
-    <a class="btn btn-light" href="#create-task">创建新任务</a>
-    <a class="btn btn-outline-light" href="#task-history">查看 Demo 数据</a>
-  </div>
 </section>
 
-<section class="card mb-4" id="create-task">
-  <div class="card-header">创建抓取任务</div>
+<section class="tool-card mb-4" id="create-task">
+  <div class="section-heading">
+    <div>
+      <p class="section-kicker">采集配置</p>
+      <h2>创建抓取任务</h2>
+    </div>
+    <span class="text-muted small">默认规模 5 × 5 × 50</span>
+  </div>
   <div class="card-body">
     <div id="form-error" class="alert alert-danger d-none" role="alert"></div>
     <form id="task-form" onsubmit="submitTask(event)">
@@ -50,13 +52,16 @@
   </div>
 </section>
 
-<section class="card" id="task-history">
-  <div class="card-header d-flex justify-content-between align-items-center">
-    <span>最近任务与 Demo 数据</span>
-    <button class="btn btn-outline-secondary btn-sm" onclick="window.location.href='/'">手动刷新</button>
+<section class="tool-card" id="task-history">
+  <div class="section-heading">
+    <div>
+      <p class="section-kicker">任务历史</p>
+      <h2>最近任务</h2>
+    </div>
+    {% if has_running_tasks %}<span class="badge text-bg-warning">自动更新中</span>{% endif %}
   </div>
   <div class="table-responsive">
-    <table class="table table-hover align-middle mb-0">
+    <table class="table app-table align-middle mb-0">
       <thead>
         <tr>
           <th>任务 ID</th>

app/templates/items/detail.html

@@ -3,27 +3,14 @@
 {% block breadcrumbs %}
 <nav aria-label="breadcrumb"><ol class="breadcrumb">
   <li class="breadcrumb-item"><a href="/">首页</a></li>
-  <li class="breadcrumb-item"><a href="/tasks/{{ task.id }}">任务 #{{ task.id }}</a></li>
-  <li class="breadcrumb-item"><a href="/hotspots/{{ hotspot.id }}/report">热点 #{{ hotspot.rank }}：{{ hotspot.title }}</a></li>
+  <li class="breadcrumb-item"><a href="/tasks/{{ task.id }}">任务 {{ task.display_id }}</a></li>
+  <li class="breadcrumb-item"><a href="/hotspots/{{ hotspot.id }}/report">热点 {{ hotspot.rank }}：{{ hotspot.title }}</a></li>
   <li class="breadcrumb-item active">{{ item.title or item.source_item_id }}</li>
 </ol></nav>
 {% endblock %}
 {% block content %}
 <div class="d-flex justify-content-between align-items-center mb-3">
   <h1 class="h3 mb-0">{{ item.title or item.source_item_id }}</h1>
-  <div class="d-flex gap-2">
-    <button
-      class="btn btn-outline-primary"
-      onclick="downloadExport('/api/export/items/{{ item.id }}.md', 'item-report.md', event)"
-      {% if not report %}disabled title="报告尚未生成"{% endif %}>
-      导出 Markdown 报告
-    </button>
-    <button
-      class="btn btn-outline-secondary"
-      onclick="downloadExport('/api/export/items/{{ item.id }}/comments.csv', 'item-comments.csv', event)">
-      导出评论 CSV
-    </button>
-  </div>
 </div>
 {% if report %}
   {% include "partials/report_panel.html" %}
@@ -31,7 +18,6 @@
   <div class="text-center text-muted py-5">
     <div class="spinner-border text-warning mb-3" role="status"></div>
     <p>报告生成中，请稍候...</p>
-    <p class="small mb-0">Markdown 报告将在分析完成后开放下载；评论 CSV 可先导出已抓取的数据。</p>
   </div>
 {% endif %}
 <section class="card"><div class="card-header">评论明细</div><div class="table-responsive">

app/templates/partials/task_rows.html

@@ -4,14 +4,11 @@
   {% set ai_percent = rate_percent(task.analysis_success_rate) %}
   <tr>
     <td>
-      <code>{{ task.id }}</code>
-      {% if task.is_demo %}
-        <br><span class="badge text-bg-info">Demo 数据</span>
-      {% endif %}
+      <span class="task-number">{{ task.display_id or loop.index }}</span>
     </td>
     <td>{{ task.platform | platform_label }}</td>
-    <td>{{ task.created_at }}</td>
-    <td><small class="text-muted">热点 {{ task.hotspot_limit }} / 内容 {{ task.item_limit_per_hotspot }} / 评论 {{ task.comment_limit_per_item }}</small></td>
+    <td>{{ task.created_at_label }}</td>
+    <td><small class="text-muted">{{ task.scale_label }}</small></td>
     <td class="task-progress-cell">
       <div class="d-flex justify-content-between align-items-center gap-2">
         <span class="small">已处理 {{ task.processed_items_count }} / 共 {{ task.total_items_count }} 条内容</span>
@@ -32,14 +29,11 @@
     </td>
     <td>
       <span class="badge {{ status_class }}">{{ status_label }}</span>
-      {% if task.error_stage or task.error_type %}
-        <br><small class="text-muted">{{ task.error_stage }} / {{ task.error_type }}</small>
-      {% endif %}
-      {% if task.error_message %}
-        <br><small class="text-danger">{{ task.error_message }}</small>
+      {% if task.error_summary %}
+        <br><small class="text-danger">{{ task.error_summary }}</small>
       {% endif %}
     </td>
-    <td><a class="btn btn-sm btn-outline-primary" href="/tasks/{{ task.id }}">查看</a></td>
+    <td><a class="btn btn-sm btn-primary" href="/tasks/{{ task.id }}">查看</a></td>
   </tr>
 {% else %}
   <tr>

app/templates/tasks/detail.html

@@ -1,24 +1,27 @@
 {% extends "base.html" %}
-{% block title %}任务 #{{ task.id }} - 热榜评论分析工具{% endblock %}
+{% block title %}任务 {{ task.display_id }} - 热榜评论分析工具{% endblock %}
 {% block breadcrumbs %}
 <nav aria-label="breadcrumb"><ol class="breadcrumb">
   <li class="breadcrumb-item"><a href="/">首页</a></li>
-  <li class="breadcrumb-item active">任务 #{{ task.id }}</li>
+  <li class="breadcrumb-item active">任务 {{ task.display_id }}</li>
 </ol></nav>
 {% endblock %}
 {% block content %}
-<div class="d-flex justify-content-between align-items-center mb-3">
-  <h1 class="h3 mb-0">任务 #{{ task.id }}</h1>
-  <button class="btn btn-outline-secondary btn-sm" onclick="window.location.reload()">手动刷新</button>
-</div>
-<section class="card mb-4" {% if task.status == "running" %}data-task-id="{{ task.id }}" data-auto-poll="true"{% endif %}><div class="card-body">
+<div class="page-heading mb-3">
+  <div>
+    <p class="section-kicker">任务详情</p>
+    <h1>任务 {{ task.display_id }}</h1>
+  </div>
   {% set label, cls = status_badge_config(task.status) %}
+  <span class="badge {{ cls }}">{{ label }}</span>
+</div>
+<section class="tool-card task-overview mb-4" {% if task.status == "running" %}data-task-id="{{ task.id }}" data-auto-poll="true"{% endif %}><div class="card-body">
   {% set percent = progress_percent(task.processed_items_count, task.total_items_count) %}
   {% set ai_percent = rate_percent(task.analysis_success_rate) %}
   {% set target_comments = task.hotspot_limit * task.item_limit_per_hotspot * task.comment_limit_per_item %}
   <div class="d-flex flex-wrap justify-content-between gap-3 mb-3">
-    <p class="mb-0">平台：{{ task.platform | platform_label }} <span class="badge {{ cls }}">{{ label }}</span>{% if task.is_demo %}<span class="badge text-bg-info ms-1">Demo 数据</span>{% endif %}</p>
-    <p class="mb-0 text-muted">阶段：{{ task.current_stage_label or "等待启动" }}</p>
+    <p class="mb-0">平台：{{ task.platform | platform_label }}</p>
+    <p class="mb-0 text-muted">创建时间 {{ task.created_at_label }} · 阶段 {{ task.current_stage_label or "等待启动" }}</p>
   </div>
   <div class="mb-3">
     <div class="d-flex justify-content-between align-items-center gap-2">
@@ -72,17 +75,17 @@
       <span class="badge text-bg-warning ms-1">AI 样本不足</span>
     {% endif %}
   </div>
-  {% if task.error_message %}<div class="alert alert-danger">{{ task.error_stage }} / {{ task.error_type }}：{{ task.error_message }}</div>{% endif %}
+  {% if task.error_summary %}<div class="alert alert-danger">{{ task.error_summary }}</div>{% endif %}
 </div></section>
 {% if task.status == "running" and not hotspots %}
 <div class="text-center text-muted py-5"><div class="spinner-border text-warning mb-3"></div><p>正在抓取热点数据，请稍候...</p></div>
 {% endif %}
 <div class="accordion" id="hotspot-list">
 {% for hotspot in hotspots %}
-  <div class="accordion-item">
+  <div class="accordion-item hotspot-item">
     <h2 class="accordion-header">
       <button class="accordion-button {% if not loop.first %}collapsed{% endif %}" type="button" data-bs-toggle="collapse" data-bs-target="#hotspot-{{ hotspot.id }}">
-        热点 #{{ hotspot.rank }}：{{ hotspot.title }}
+        热点 {{ hotspot.rank }}：{{ hotspot.title }}
       </button>
     </h2>
     <div id="hotspot-{{ hotspot.id }}" class="accordion-collapse collapse {% if loop.first %}show{% endif %}" data-bs-parent="#hotspot-list">

docs/Deployment.md

@@ -99,9 +99,21 @@ docker compose up -d --build
 curl -f http://localhost:8000/health
 ```
 
-## 公网云服务器部署
+## 公网部署决策清单
 
-第一版公网演示使用云服务器 + Docker Compose，不增加登录或密码。
+上线前必须确认以下决策，未经确认不要直接开放公网：
+
+1. 部署方式：云服务器 Docker Compose、PaaS 平台，还是临时内网穿透演示。
+2. 访问控制：是否需要访问密码 / 简单登录，或只在可信网络内演示。
+3. 成本控制：公网用户是否允许直接消耗真实 TikHub 和 AI Key。
+4. Demo 数据：公网环境使用脱敏 Demo 数据，还是允许展示真实抓取结果。
+5. 数据生命周期：SQLite 数据是否需要持久保留，以及如何备份 / 重置。
+
+当前建议：在上述问题确认前，只做本地 Docker 验收和部署准备，不把端口直接暴露到公网。
+
+## 公网云服务器部署候选方案
+
+如果确认采用云服务器 Docker Compose，可按以下步骤执行。
 
 上线前准备：
 

docs/MVP-WorkOrders.md

@@ -1269,6 +1269,35 @@ chore: 增加公网部署配置
 - 是否允许公网用户直接消耗真实 TikHub 和 AI Key。
 - 是否需要限制同一时间只能运行一个任务。
 
+当前推进记录：
+
+- 日期：2026-07-03
+- 相关改动：
+  - `docs/Deployment.md` 新增公网部署决策清单，明确上线前必须确认部署方式、访问密码、真实 TikHub / AI Key 消耗、Demo 数据和数据生命周期。
+  - 将“默认云服务器 Docker Compose 且不加密码”的表述改为候选方案，避免未经确认直接开放公网。
+  - `docs/UserGuide.md` 补充公网云服务器候选部署步骤、Demo 数据初始化和数据库异常备份路径。
+  - 补充文档测试，锁定部署文档必须包含上线前确认项。
+- 验证命令：
+  - `docker exec hot-comments-tool-app-1 sh -lc 'python -m pytest /app/tests/unit/test_docs.py -q'`
+  - `curl -f http://localhost:8000/health`
+  - `curl -s -o /tmp/wo15-index.html -w '%{http_code}' http://localhost:8000/`
+  - `docker compose ps`
+  - `docker compose up -d --build`
+- 验证结果：
+  - 文档测试：`2 passed`
+  - 本地 health：返回 `{"status":"ok"}`
+  - 首页：HTTP 200
+  - 当前运行容器：`hot-comments-tool-app-1`，端口 `0.0.0.0:8000->8000/tcp`
+  - 镜像重建：仍卡在 `pip install uv` 下载阶段，手动中断，未完成新镜像 rebuild。
+- 当前结论：
+  - 本地服务可访问，部署文档已进入更安全的“待确认后上线”状态。
+  - 未执行公网部署，原因是 WO-15 的关键产品 / 安全决策仍待用户确认，且当前网络环境下镜像重建仍不稳定。
+- 后续必须确认：
+  - 部署方式：云服务器 Docker Compose / PaaS / 临时内网穿透。
+  - 是否需要访问密码 / 简单登录。
+  - 是否允许公网访问者直接创建真实抓取任务并消耗 TikHub / AI Key。
+  - 是否先完成可复现 Demo 数据方案（WO-14）再公网演示。
+
 ### WO-16 UI 信息简化与产品化文案清理
 
 优先级：P1
@@ -1325,6 +1354,29 @@ feat: 优化 MVP 页面产品化体验
 - UI 风格方向：更偏数据仪表盘、内部工具，还是偏演示型产品页面。
 - 是否需要提供简单品牌名 / Logo / 说明文案。
 
+完成记录：
+
+- 完成日期：2026-07-03
+- 相关改动：
+  - 任务列表和任务详情默认展示顺序号 `#1/#2/#3`，不再把完整 UUID 铺在列表和标题中。
+  - 任务详情页保留“完整 ID”排查信息，方便复制给开发定位。
+  - 创建时间压缩为 `YYYY-MM-DD HH:mm`，配置规模压缩为 `小红书 · 5热点 × 5内容 × 50评论`。
+  - 失败状态只展示中文错误说明，不再直接展示 `system / unexpected_restart`、`recover / interrupted` 等内部枚举。
+  - 移除页脚 `内部演示工具 | 仅供学习参考`。
+- 验证命令：
+  - `docker cp app/. hot-comments-tool-app-1:/app/app/`
+  - `docker cp tests/. hot-comments-tool-app-1:/app/tests/`
+  - `docker exec hot-comments-tool-app-1 sh -lc 'python -m pytest /app/tests/integration/test_routes.py::test_index_page_uses_short_task_numbers_compact_fields_and_friendly_error_copy /app/tests/integration/test_routes.py::test_task_detail_uses_short_number_title_and_keeps_full_uuid_for_diagnostics -q'`
+  - `docker exec hot-comments-tool-app-1 sh -lc 'python -m pytest /app/tests/integration/test_routes.py -q'`
+- 验证结果：
+  - WO-16 聚焦测试：`2 passed, 1 warning`
+  - 路由集成测试：`23 passed, 1 warning`
+- 验收结论：
+  - 任务列表、任务详情、报告页和内容详情页的任务引用已改为短编号。
+  - 用户界面不再展示内部英文错误枚举和演示页脚文案。
+- 遗留问题：
+  - 本轮不做大规模视觉重设计，品牌名 / Logo 仍需用户确认后再纳入正式工单。
+
 ### WO-17 默认规模数据量解释与展示优化
 
 优先级：P1

docs/UserGuide.md

@@ -93,16 +93,13 @@ http://localhost:8000
 
 如果报告尚未生成，页面会显示“报告生成中，请稍候...”，不会返回 500。
 
-## 5. 导出文件
+## 5. 文件导出
 
-支持导出：
+当前版本已关闭 CSV / Markdown 文件下载入口和接口。页面仍可直接查看：
 
-- 热点报告 Markdown
-- 内容报告 Markdown
-- 热点评论 CSV
-- 内容评论 CSV
-
-CSV 使用 UTF-8-BOM，适合 Windows Excel 打开。评论内容中的换行会被替换为空格，`= + - @` 开头的内容会自动加单引号，避免被 Excel 当作公式执行。
+- 热点级汇总报告
+- 内容条目级报告
+- 评论明细
 
 ## 6. 小规模验收
 
@@ -119,7 +116,7 @@ CSV 使用 UTF-8-BOM，适合 Windows Excel 打开。评论内容中的换行会
 - 至少有热点、内容、评论、报告。
 - AI 成功率正常，或页面有明确“AI 样本不足”提示。
 - 能打开热点报告、内容详情页。
-- 能导出 Markdown 和 CSV。
+- 页面不展示 CSV / Markdown 下载入口。
 
 ## 7. 默认规模验收
 
@@ -197,6 +194,31 @@ docker compose up -d --build
 curl -f http://localhost:8000/health
 ```
 
+### 公网云服务器部署
+
+公网部署前先确认部署方式、访问密码、是否允许消耗真实 TikHub 和 AI Key。未经确认不要直接开放公网。
+
+如果确认使用云服务器 Docker Compose，步骤与本地类似：
+
+```bash
+git pull
+cp .env.example .env
+docker compose up -d --build
+curl -f http://127.0.0.1:8000/health
+```
+
+公网演示建议先初始化脱敏 Demo 数据：
+
+```bash
+docker compose exec app python -m app.demo_seed
+```
+
+如果页面出现数据库不可用提示，系统会尽量把 SQLite 文件备份到：
+
+```text
+data/corrupt-backups
+```
+
 ## 9. 推荐提交节奏
 
 按 `docs/MVP-WorkOrders.md` 的工单推进：

requirements.txt

@@ -0,0 +1,8 @@
+beautifulsoup4>=4.12.3
+fastapi>=0.115.0
+httpx>=0.27.0
+jinja2>=3.1.4
+pydantic>=2.8.0
+pydantic-settings>=2.4.0
+sqlalchemy>=2.0.32
+uvicorn[standard]>=0.30.6

tests/integration/test_routes.py

@@ -1,5 +1,7 @@
 from datetime import UTC, datetime, timedelta
 
+from bs4 import BeautifulSoup
+
 from app.models import Comment, ContentItem, Hotspot, Report, Task
 from tests.helpers import make_test_client
 from sqlalchemy.exc import OperationalError
@@ -16,7 +18,10 @@ def test_index_page_renders_task_form_empty_state_and_default_scale():
     assert 'name="item_limit_per_hotspot"' in response.text
     assert 'name="comment_limit_per_item"' in response.text
     assert "1250" in response.text
-    assert 'onclick="window.location.href=\'/\'"' in response.text
+    assert "手动刷新" not in response.text
+    assert 'onclick="window.location.href=\'/\'"' not in response.text
+    assert 'href="#create-task"' not in response.text
+    assert 'href="#task-history"' not in response.text
     assert "还没有任何任务" in response.text
 
 
@@ -59,6 +64,62 @@ def test_index_page_lists_existing_tasks():
     assert "DOMContentLoaded" in response.text
 
 
+def test_index_page_uses_short_task_numbers_compact_fields_and_friendly_error_copy():
+    first_id = "11111111-1111-4111-8111-111111111111"
+    second_id = "22222222-2222-4222-8222-222222222222"
+    with make_test_client() as (client, engine):
+        from sqlalchemy.orm import Session
+
+        with Session(engine) as session:
+            session.add(
+                Task(
+                    id=first_id,
+                    platform="xiaohongshu",
+                    status="failed",
+                    created_at=datetime(2026, 7, 3, 10, 30, tzinfo=UTC),
+                    hotspot_limit=5,
+                    item_limit_per_hotspot=5,
+                    comment_limit_per_item=50,
+                    error_stage="system",
+                    error_type="unexpected_restart",
+                    error_message="系统重启，任务被中断",
+                )
+            )
+            session.add(
+                Task(
+                    id=second_id,
+                    platform="douyin",
+                    status="success",
+                    created_at=datetime(2026, 7, 3, 10, 35, tzinfo=UTC),
+                    hotspot_limit=1,
+                    item_limit_per_hotspot=1,
+                    comment_limit_per_item=10,
+                )
+            )
+            session.commit()
+
+        response = client.get("/")
+
+    assert response.status_code == 200
+    visible_text = BeautifulSoup(response.text, "html.parser").get_text(" ")
+    assert " 1 " in f" {visible_text} "
+    assert " 2 " in f" {visible_text} "
+    assert "#" not in visible_text
+    assert first_id not in visible_text
+    assert second_id not in visible_text
+    assert "2026-07-03 18:30" in visible_text
+    assert "2026-07-03 18:35" in visible_text
+    assert "Demo 数据" not in visible_text
+    assert "5热点 × 5内容 × 50评论" in visible_text
+    assert "1热点 × 1内容 × 10评论" in visible_text
+    assert "小红书 · 5热点" not in visible_text
+    assert "抖音 · 1热点" not in visible_text
+    assert "系统重启，任务被中断" in visible_text
+    assert "system / unexpected_restart" not in visible_text
+    assert "unexpected_restart" not in visible_text
+    assert "内部演示工具 | 仅供学习参考" not in visible_text
+
+
 def test_index_page_does_not_auto_poll_without_running_tasks():
     with make_test_client() as (client, engine):
         from sqlalchemy.orm import Session
@@ -112,11 +173,75 @@ def test_running_task_detail_page_auto_polls_current_task():
     assert response.status_code == 200
     assert 'data-task-id="task-running"' in response.text
     assert "pollTaskDetailStatus" in response.text
-    assert 'onclick="window.location.reload()"' in response.text
+    assert "手动刷新" not in response.text
+    assert 'onclick="window.location.reload()"' not in response.text
     assert "已处理 1 / 共 2 条内容" in response.text
     assert "AI 成功率 100%" in response.text
 
 
+def test_task_detail_uses_short_number_title_and_hides_full_uuid():
+    task_id = "33333333-3333-4333-8333-333333333333"
+    with make_test_client() as (client, engine):
+        from sqlalchemy.orm import Session
+
+        with Session(engine) as session:
+            session.add(
+                Task(
+                    id=task_id,
+                    platform="xiaohongshu",
+                    status="running",
+                    created_at=datetime(2026, 7, 3, 10, 30, tzinfo=UTC),
+                    hotspot_limit=5,
+                    item_limit_per_hotspot=5,
+                    comment_limit_per_item=50,
+                )
+            )
+            session.commit()
+
+        response = client.get(f"/tasks/{task_id}")
+
+    assert response.status_code == 200
+    visible_text = BeautifulSoup(response.text, "html.parser").get_text(" ")
+    assert "<title>任务 1 - 热榜评论分析工具</title>" in response.text
+    assert "任务 1" in visible_text
+    assert "完整 ID" not in visible_text
+    assert task_id not in visible_text
+    assert "任务 #33333333-3333-4333-8333-333333333333" not in visible_text
+    assert "#" not in visible_text
+    assert "创建时间 2026-07-03 18:30" in visible_text
+    assert "Demo 数据" not in visible_text
+
+
+def test_demo_task_flag_is_kept_but_demo_copy_is_hidden_from_pages():
+    with make_test_client() as (client, engine):
+        from sqlalchemy.orm import Session
+
+        with Session(engine) as session:
+            session.add(
+                Task(
+                    id="demo-task-hidden-copy",
+                    platform="xiaohongshu",
+                    status="success",
+                    created_at=datetime(2026, 7, 3, 10, 30, tzinfo=UTC),
+                    hotspot_limit=1,
+                    item_limit_per_hotspot=1,
+                    comment_limit_per_item=10,
+                )
+            )
+            session.commit()
+
+        index_response = client.get("/")
+        detail_response = client.get("/tasks/demo-task-hidden-copy")
+        api_response = client.get("/api/tasks/demo-task-hidden-copy")
+
+    assert index_response.status_code == 200
+    assert detail_response.status_code == 200
+    assert api_response.status_code == 200
+    assert api_response.json()["is_demo"] is True
+    assert "Demo 数据" not in BeautifulSoup(index_response.text, "html.parser").get_text(" ")
+    assert "Demo 数据" not in BeautifulSoup(detail_response.text, "html.parser").get_text(" ")
+
+
 def test_running_task_detail_page_keeps_polling_after_hotspots_exist():
     with make_test_client() as (client, engine):
         from sqlalchemy.orm import Session
@@ -209,7 +334,7 @@ def test_running_task_detail_page_shows_stage_runtime_and_stale_progress_warning
         response = client.get("/tasks/task-stale")
 
     assert response.status_code == 200
-    assert "阶段：AI 分析中" in response.text
+    assert "阶段 AI 分析中" in response.text
     assert "运行时长" in response.text
     assert "2小时5分钟" in response.text
     assert "最近进度" in response.text
@@ -241,7 +366,8 @@ def test_failed_task_detail_page_shows_failure_reason_without_loading_copy():
 
     assert response.status_code == 200
     assert "失败" in response.text
-    assert "recover / interrupted：系统重启，任务被中断" in response.text
+    assert "系统重启，任务被中断" in response.text
+    assert "recover / interrupted" not in response.text
     assert "正在抓取热点数据" not in response.text
     assert "pollTaskDetailStatus" not in response.text
 
@@ -330,11 +456,15 @@ def test_result_pages_render_seeded_data():
 
     assert task_response.status_code == 200
     assert "热点标题" in task_response.text
-    assert 'onclick="window.location.reload()"' in task_response.text
+    assert "热点 1：热点标题" in BeautifulSoup(task_response.text, "html.parser").get_text(" ")
+    assert 'onclick="window.location.reload()"' not in task_response.text
+    assert "手动刷新" not in task_response.text
     assert hotspot_response.status_code == 200
     assert "热点总结" in hotspot_response.text
-    assert "downloadExport('/api/export/hotspots/hot-result.md'" in hotspot_response.text
-    assert "downloadExport('/api/export/hotspots/hot-result/comments.csv'" in hotspot_response.text
+    assert "downloadExport(" not in hotspot_response.text
+    assert "/api/export/" not in hotspot_response.text
+    assert "导出 Markdown" not in hotspot_response.text
+    assert "导出热点评论 CSV" not in hotspot_response.text
     assert "评论样本" in hotspot_response.text
     assert "关联内容" in hotspot_response.text
     assert "正向" in hotspot_response.text
@@ -345,8 +475,13 @@ def test_result_pages_render_seeded_data():
     assert "<pre" not in hotspot_response.text
     assert item_response.status_code == 200
     assert "内容总结" in item_response.text
-    assert "downloadExport('/api/export/items/item-result.md'" in item_response.text
-    assert "downloadExport('/api/export/items/item-result/comments.csv'" in item_response.text
+    item_visible_text = BeautifulSoup(item_response.text, "html.parser").get_text(" ")
+    assert "热点 1：热点标题" in item_visible_text
+    assert "#" not in item_visible_text
+    assert "downloadExport(" not in item_response.text
+    assert "/api/export/" not in item_response.text
+    assert "导出 Markdown" not in item_response.text
+    assert "导出评论 CSV" not in item_response.text
     assert "评论样本" in item_response.text
     assert "Top 标签" in item_response.text
     assert "评论内容" in item_response.text
@@ -381,14 +516,16 @@ def test_report_pages_render_empty_state_when_report_missing():
 
     assert hotspot_response.status_code == 200
     assert "报告生成中，请稍候" in hotspot_response.text
-    assert 'title="报告尚未生成"' in hotspot_response.text
-    assert "Markdown 报告将在分析完成后开放下载" in hotspot_response.text
-    assert "downloadExport('/api/export/hotspots/hot-no-report/comments.csv'" in hotspot_response.text
+    assert 'title="报告尚未生成"' not in hotspot_response.text
+    assert "Markdown 报告将在分析完成后开放下载" not in hotspot_response.text
+    assert "downloadExport(" not in hotspot_response.text
+    assert "/api/export/" not in hotspot_response.text
     assert item_response.status_code == 200
     assert "报告生成中，请稍候" in item_response.text
-    assert 'title="报告尚未生成"' in item_response.text
-    assert "Markdown 报告将在分析完成后开放下载" in item_response.text
-    assert "downloadExport('/api/export/items/item-no-report/comments.csv'" in item_response.text
+    assert 'title="报告尚未生成"' not in item_response.text
+    assert "Markdown 报告将在分析完成后开放下载" not in item_response.text
+    assert "downloadExport(" not in item_response.text
+    assert "/api/export/" not in item_response.text
 
 
 def test_report_pages_warn_when_ai_sample_is_insufficient():
@@ -411,7 +548,7 @@ def test_report_pages_warn_when_ai_sample_is_insufficient():
     assert "当前有效评论样本不足" in item_response.text
 
 
-def test_export_routes_return_csv_and_markdown():
+def test_export_routes_are_removed():
     with make_test_client() as (client, engine):
         seed_result_data(engine)
 
@@ -420,15 +557,10 @@ def test_export_routes_return_csv_and_markdown():
         hotspot_md = client.get("/api/export/hotspots/hot-result.md")
         item_md = client.get("/api/export/items/item-result.md")
 
-    assert csv_response.status_code == 200
-    assert csv_response.content.startswith(b"\xef\xbb\xbf")
-    assert "评论内容".encode("utf-8") in csv_response.content
-    assert hotspot_csv_response.status_code == 200
-    assert "评论内容".encode("utf-8") in hotspot_csv_response.content
-    assert hotspot_md.status_code == 200
-    assert "热点总结" in hotspot_md.text
-    assert item_md.status_code == 200
-    assert "内容总结" in item_md.text
+    assert csv_response.status_code == 404
+    assert hotspot_csv_response.status_code == 404
+    assert hotspot_md.status_code == 404
+    assert item_md.status_code == 404
 
 
 def test_api_tasks_returns_service_unavailable_when_database_has_io_error(monkeypatch):
@@ -524,18 +656,6 @@ def test_item_detail_page_shows_database_error_state_when_database_has_io_error(
     assert "数据库暂时不可用" in response.text
 
 
-def test_export_route_returns_service_unavailable_when_database_has_io_error(monkeypatch):
-    def broken_export(_session, _item_id):
-        raise OperationalError("SELECT 1", {}, Exception("disk I/O error"))
-
-    monkeypatch.setattr("app.main.export_item_comments_csv", broken_export)
-
-    with make_test_client() as (client, _engine):
-        response = client.get("/api/export/items/item-io/comments.csv")
-
-    assert response.status_code == 503
-    assert response.json()["detail"] == "数据库暂时不可用，请稍后重试或联系维护者恢复数据。"
-
 
 def test_missing_html_pages_render_friendly_not_found_page():
     with make_test_client() as (client, _engine):

tests/integration/test_task_creation.py

@@ -1,5 +1,8 @@
+from datetime import UTC, datetime, timedelta
+
 from fastapi.testclient import TestClient
 from sqlalchemy import create_engine
+from sqlalchemy.orm import Session
 from sqlalchemy.pool import StaticPool
 
 from app.db import Base, get_db_session
@@ -92,6 +95,47 @@ def test_create_task_returns_400_when_running_task_exists():
         app.dependency_overrides.clear()
 
 
+def test_create_task_recovers_stale_running_task_before_creating_new_one(monkeypatch):
+    client, engine = make_test_client()
+    now = datetime(2026, 7, 3, 12, 0, tzinfo=UTC)
+    monkeypatch.setattr("app.services.task_service.utc_now", lambda: now)
+    monkeypatch.setattr("app.services.task_service.task_executor.submit", lambda *_args, **_kwargs: None)
+    try:
+        with Session(engine) as session:
+            session.add(
+                Task(
+                    id="stale-running",
+                    platform="douyin",
+                    status="running",
+                    current_stage="ai_analysis",
+                    last_progress_at=now - timedelta(minutes=11),
+                )
+            )
+            session.commit()
+
+        response = client.post(
+            "/api/tasks",
+            json={
+                "platform": "douyin",
+                "hotspot_limit": 5,
+                "item_limit_per_hotspot": 5,
+                "comment_limit_per_item": 50,
+            },
+        )
+
+        with Session(engine) as session:
+            stale_task = session.get(Task, "stale-running")
+
+        assert response.status_code == 201
+        assert stale_task.status == "failed"
+        assert stale_task.error_stage == "system"
+        assert stale_task.error_type == "stale_progress_timeout"
+        assert "超过 10 分钟没有进度更新" in stale_task.error_message
+    finally:
+        engine.dispose()
+        app.dependency_overrides.clear()
+
+
 def test_task_list_and_detail_return_created_tasks():
     client, engine = make_test_client()
     try:

tests/integration/test_task_flow_xiaohongshu.py

@@ -1,4 +1,6 @@
 import json
+import time
+from types import SimpleNamespace
 
 from sqlalchemy import select
 from sqlalchemy.orm import Session
@@ -147,3 +149,43 @@ def test_xiaohongshu_task_flow_keeps_item_success_when_ai_request_raises(monkeyp
             assert persisted.analysis_success_rate == 0.0
             assert comment.ai_analysis_status == "failed"
             assert comment.reason == "ai_parse_failed"
+
+
+def test_xiaohongshu_task_flow_times_out_stalled_ai_request(monkeypatch):
+    with make_test_client() as (_client, engine):
+        monkeypatch.setattr("app.services.task_service.build_platform", lambda _platform: FakeXhsPlatform())
+        monkeypatch.setattr(
+            "app.services.task_service.get_settings",
+            lambda: SimpleNamespace(ai_max_retries=1, ai_timeout_seconds=0.01),
+        )
+
+        def stalled_requester(prompt):
+            time.sleep(0.2)
+            comments = json.loads(prompt[prompt.index("[") :])
+            return json.dumps(
+                [{"comment_id": comments[0]["comment_id"], "sentiment": "positive", "labels": ["超时后不应采用"], "reason": "late"}],
+                ensure_ascii=False,
+            )
+
+        monkeypatch.setattr("app.services.task_service.build_ai_dependencies", lambda: (stalled_requester, None))
+
+        with Session(engine) as session:
+            task = create_task(
+                session,
+                CreateTaskRequest(platform="xiaohongshu", hotspot_limit=1, item_limit_per_hotspot=1, comment_limit_per_item=10),
+                submit_background=False,
+            )
+            task_id = task.id
+
+        with Session(engine) as session:
+            run_task(task_id, session_factory=lambda: session)
+
+            persisted = session.get(Task, task_id)
+            comment = session.scalar(select(Comment).where(Comment.task_id == task_id))
+
+        assert persisted.status == "success"
+        assert persisted.finished_at is not None
+        assert persisted.analysis_status == "insufficient"
+        assert comment.ai_analysis_status == "failed"
+        assert comment.sentiment == "unknown"
+        assert comment.reason == "ai_parse_failed"

tests/unit/test_comment_pagination.py

@@ -51,3 +51,17 @@ def test_tikhub_client_raises_structured_error_after_retries(monkeypatch):
     assert exc_info.value.status_code == 429
     assert "secret-token" not in str(exc_info.value)
     assert sleeps == [1, 2, 4]
+
+
+def test_tikhub_client_reports_401_as_auth_error_without_leaking_token():
+    transport = SequenceTransport([httpx.Response(401, json={"message": "Unauthorized"})])
+    http_client = httpx.Client(transport=httpx.MockTransport(transport))
+    client = TikHubClient(base_url="https://api.test", api_key="secret-token", http_client=http_client)
+
+    with pytest.raises(PlatformAPIError) as exc_info:
+        client.get("/demo")
+
+    assert exc_info.value.error_type == "auth_error"
+    assert exc_info.value.status_code == 401
+    assert str(exc_info.value) == "TikHub 鉴权失败，请检查 TIKHUB_API_KEY 是否有效"
+    assert "secret-token" not in str(exc_info.value)

tests/unit/test_db_stability.py

@@ -1,4 +1,5 @@
 import pytest
+from sqlalchemy.pool import NullPool
 
 from app.db import check_database_integrity, checkpoint_sqlite_wal, create_sqlite_engine, ensure_sqlite_schema_compat
 
@@ -11,6 +12,14 @@ def test_check_database_integrity_returns_ok_for_valid_sqlite_database():
         engine.dispose()
 
 
+def test_file_sqlite_engine_does_not_reuse_connections_after_operational_errors(tmp_path):
+    engine = create_sqlite_engine(f"sqlite:///{tmp_path / 'app.db'}")
+    try:
+        assert isinstance(engine.pool, NullPool)
+    finally:
+        engine.dispose()
+
+
 def test_check_database_integrity_raises_when_sqlite_reports_problem(monkeypatch):
     class FakeCursor:
         def execute(self, _sql):

tests/unit/test_deployment_config.py

@@ -29,3 +29,19 @@ def test_real_env_file_is_gitignored():
     gitignore_lines = (ROOT / ".gitignore").read_text(encoding="utf-8").splitlines()
 
     assert ".env" in gitignore_lines
+
+
+def test_dockerfile_does_not_bootstrap_uv_from_runtime_pip():
+    dockerfile_content = (ROOT / "Dockerfile").read_text(encoding="utf-8")
+
+    assert "pip install --no-cache-dir uv" not in dockerfile_content
+    assert "uv pip install" not in dockerfile_content
+
+
+def test_dockerfile_installs_project_without_build_isolation():
+    dockerfile_content = (ROOT / "Dockerfile").read_text(encoding="utf-8")
+
+    assert "COPY requirements.txt ./" in dockerfile_content
+    assert "pip install --no-cache-dir -r requirements.txt" in dockerfile_content
+    assert "pip install --no-cache-dir ." not in dockerfile_content
+    assert "pip install --no-cache-dir --no-build-isolation ." not in dockerfile_content

tests/unit/test_docs.py

@@ -4,7 +4,7 @@ from pathlib import Path
 ROOT = Path(__file__).resolve().parents[2]
 
 
-def test_user_guide_covers_startup_acceptance_exports_and_troubleshooting():
+def test_user_guide_covers_startup_acceptance_no_exports_and_troubleshooting():
     content = (ROOT / "docs" / "UserGuide.md").read_text(encoding="utf-8")
 
     for required in [
@@ -12,7 +12,9 @@ def test_user_guide_covers_startup_acceptance_exports_and_troubleshooting():
         "http://localhost:8000",
         "1 × 1 × 10",
         "5 × 5 × 50",
-        "导出文件",
+        "文件导出",
+        "已关闭 CSV / Markdown 文件下载入口和接口",
+        "页面不展示 CSV / Markdown 下载入口",
         "8000 端口被占用",
         "API Key 缺失",
         "任务长期 running",
@@ -22,3 +24,18 @@ def test_user_guide_covers_startup_acceptance_exports_and_troubleshooting():
         "data/corrupt-backups",
     ]:
         assert required in content
+
+
+def test_deployment_doc_requires_public_access_decisions_before_going_online():
+    content = (ROOT / "docs" / "Deployment.md").read_text(encoding="utf-8")
+
+    for required in [
+        "上线前必须确认",
+        "部署方式",
+        "访问密码",
+        "真实 TikHub 和 AI Key",
+        "未经确认不要直接开放公网",
+    ]:
+        assert required in content
+
+    assert "第一版公网演示使用云服务器 + Docker Compose，不增加登录或密码" not in content

tests/unit/test_export.py

@@ -1,24 +0,0 @@
-from app.services.export_service import labels_to_text, safe_csv_field, safe_filename
-
-
-def test_labels_to_text_joins_json_array_with_chinese_comma():
-    assert labels_to_text('["认可", "质量好"]') == "认可，质量好"
-    assert labels_to_text("bad-json") == ""
-
-
-def test_safe_csv_field_prevents_formula_injection_and_newlines():
-    assert safe_csv_field("=1+1") == "'=1+1"
-    assert safe_csv_field("+1+1") == "'+1+1"
-    assert safe_csv_field("-1+1") == "'-1+1"
-    assert safe_csv_field("@SUM(1,1)") == "'@SUM(1,1)"
-    assert safe_csv_field("hello\nworld") == "hello world"
-    assert safe_csv_field("hello\r\nworld") == "hello world"
-    assert safe_csv_field("hello\rworld") == "hello world"
-
-
-def test_safe_filename_replaces_illegal_chars_and_truncates_keyword():
-    filename = safe_filename("douyin", "task", "a/b:c*d?e<f>g|很长很长很长很长很长")
-
-    assert "/" not in filename
-    assert "__" not in filename
-    assert filename.endswith(".csv")