From 633a350bfcd07dc59575ae6664efc57e919a7e24 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E8=91=A3=E5=AE=8F=E5=87=AF?= <hkdong@kland.com.cn>
Date: Sun, 31 May 2026 14:42:51 +0800
Subject: [PATCH] =?UTF-8?q?AI=E5=86=99=E4=BD=9C=EF=BC=9A=E6=8A=80=E8=83=BD?=
 =?UTF-8?q?=E6=A3=80=E7=B4=A2=E8=BF=9B=E5=BA=A6=E5=8F=8D=E9=A6=88=E4=B8=8E?=
 =?UTF-8?q?=E8=B6=85=E6=97=B6=E5=85=9C=E5=BA=95?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

检索等待时 SkillStepsCard 显示 spinner；单技能检索 25s 超时避免卡死；前后端增加 skill-timing 调试日志。

Co-authored-by: Cursor <cursoragent@cursor.com>
---
 .../components/ai-writing/SkillStepsCard.tsx  | 27 +++++++++++++-----
 .../open-canvas/hooks/useAIWritingStream.ts   | 13 +++++++++
 .../agents/ai_writing/nodes/researcher.py     | 28 ++++++++++++++++---
 .../agents/ai_writing/search/skill_search.py  |  8 ++++++
 4 files changed, 65 insertions(+), 11 deletions(-)

diff --git a/frontend-web/src/open-canvas/components/ai-writing/SkillStepsCard.tsx b/frontend-web/src/open-canvas/components/ai-writing/SkillStepsCard.tsx
index a243e295..cb8c349c 100644
--- a/frontend-web/src/open-canvas/components/ai-writing/SkillStepsCard.tsx
+++ b/frontend-web/src/open-canvas/components/ai-writing/SkillStepsCard.tsx
@@ -32,6 +32,14 @@ export function SkillStepsCard({ steps, inProgress }: SkillStepsCardProps) {
   const [expanded, setExpanded] = useState(true);
   if (steps.length === 0 && !inProgress) return null;
 
+  // 进行中的步骤：最后一条是 invoke_skill 且还没出对应的 skill_result，
+  // 说明该技能的 web 检索正在 await 中——把它的序号徽标换成 spinner，
+  // 让每次（可能数秒~数十秒的）检索等待都有"正在调用技能X检索中"的实时反馈。
+  const activeIdx =
+    inProgress && steps.length > 0 && steps[steps.length - 1].type === 'invoke_skill'
+      ? steps.length - 1
+      : -1;
+
   return (
     <div className="rounded-md border border-violet-200 bg-violet-50/40 dark:border-violet-900/40 dark:bg-violet-950/20 overflow-hidden">
       <button
@@ -63,19 +71,22 @@ export function SkillStepsCard({ steps, inProgress }: SkillStepsCardProps) {
             const label = STEP_LABEL[step.type] ?? step.type;
             const isStopEarly = step.type === 'stop_early';
             const isResult = step.type === 'skill_result';
+            const isActive = idx === activeIdx;
             return (
               <li key={idx} className="flex items-start gap-2 text-[11px]">
                 <span
                   className={cn(
                     'shrink-0 mt-0.5 inline-flex items-center justify-center size-4 rounded-full text-[10px] font-medium',
-                    isStopEarly
-                      ? 'bg-amber-100 dark:bg-amber-900/60 text-amber-600 dark:text-amber-300'
-                      : isResult
-                        ? 'bg-emerald-100 dark:bg-emerald-900/60 text-emerald-600 dark:text-emerald-300'
-                        : 'bg-violet-100 dark:bg-violet-900/60 text-violet-600 dark:text-violet-300',
+                    isActive
+                      ? 'bg-violet-500 text-white'
+                      : isStopEarly
+                        ? 'bg-amber-100 dark:bg-amber-900/60 text-amber-600 dark:text-amber-300'
+                        : isResult
+                          ? 'bg-emerald-100 dark:bg-emerald-900/60 text-emerald-600 dark:text-emerald-300'
+                          : 'bg-violet-100 dark:bg-violet-900/60 text-violet-600 dark:text-violet-300',
                   )}
                 >
-                  {idx + 1}
+                  {isActive ? <Loader2 className="size-2.5 animate-spin" /> : idx + 1}
                 </span>
                 <div className="flex-1 min-w-0">
                   <div
@@ -137,7 +148,9 @@ export function SkillStepsCard({ steps, inProgress }: SkillStepsCardProps) {
               ))}
             </>
           )}
-          {inProgress && steps.length > 0 && (
+          {inProgress && steps.length > 0 && activeIdx < 0 && (
+            // 当前不在某个技能的检索等待中（最后一条是 result / select 等），
+            // 提示还有后续步骤；正在检索时由该步自身的 spinner 体现进行中，不重复。
             <li className="flex items-center gap-2 text-[11px] text-muted-foreground italic">
               <Loader2 className="size-3 animate-spin shrink-0" />
               更多步骤生成中…
diff --git a/frontend-web/src/open-canvas/hooks/useAIWritingStream.ts b/frontend-web/src/open-canvas/hooks/useAIWritingStream.ts
index 94b5ada4..21d31bca 100644
--- a/frontend-web/src/open-canvas/hooks/useAIWritingStream.ts
+++ b/frontend-web/src/open-canvas/hooks/useAIWritingStream.ts
@@ -634,12 +634,25 @@ export function useAIWriting() {
       dispatch({ type: 'set_session', sessionId: meta.thread_id });
     },
     onCustomEvent(event: unknown) {
+      // 技能检索时机 trace：开 window.__AI_WRITING_DEBUG__ = true 后，控制台按
+      // performance.now() 打印每条 custom 事件到达时刻 + type，用来钉死
+      // skill_search_step（技能检索过程卡）与 skill_select_chunk（筛选技能打字）
+      // 的实际到达顺序，区分"后端未实时 emit" vs "前端渲染滞后"。
+      if (typeof window !== 'undefined' && (window as { __AI_WRITING_DEBUG__?: boolean }).__AI_WRITING_DEBUG__) {
+        const t = (event as { type?: string } | null)?.type;
+        // eslint-disable-next-line no-console
+        console.debug(`[ai-writing:custom] +${Math.round(performance.now())}ms type=${t ?? '?'}`, event);
+      }
       const pe = customToProgressEvent(event, state.sessionId ?? '');
       if (pe) dispatch({ type: 'handle', event: pe });
     },
     onUpdateEvent(data: Record<string, unknown> | undefined) {
       // data 形如 { node_name: { ...node_output } }，可能多个节点同一 step
       if (!data) return;
+      if (typeof window !== 'undefined' && (window as { __AI_WRITING_DEBUG__?: boolean }).__AI_WRITING_DEBUG__) {
+        // eslint-disable-next-line no-console
+        console.debug(`[ai-writing:updates] +${Math.round(performance.now())}ms nodes=${Object.keys(data).join(',')}`);
+      }
       // 兜底：updates 通道可能直接带 ``__interrupt__`` 数组（节点进 interrupt 时
       // LangGraph 推 ``{__interrupt__: [...]}``），SDK 主路径只在 values 通道
       // 暴露 stream.interrupt —— 万一后端 serialize 又把 __interrupt__ 滤掉，
diff --git a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/researcher.py b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/researcher.py
index 1f219e8a..711ae85f 100644
--- a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/researcher.py
+++ b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/researcher.py
@@ -4,6 +4,7 @@ from __future__ import annotations
 import asyncio
 import logging
 import re
+import time
 import uuid
 from typing import List
 
@@ -29,6 +30,11 @@ _MAX_MAX_MATERIALS = 100
 _DEFAULT_KEYWORD_COUNT = 6
 _MAX_KEYWORD_COUNT = 10
 
+# 单个技能的 web 检索超时（秒）。技能检索是顺序循环、最多 5 个技能，
+# DuckDuckGo 等 provider 偶发限速/挂起会把整段"技能检索过程"拖到几十秒看起来卡死；
+# 超时即放弃该技能、emit 0 条结果继续下一个，保证时间线持续推进。
+_SKILL_SEARCH_TIMEOUT_SECONDS = 25
+
 
 def _resolve_keyword_count(raw: object) -> int:
     """把用户传入的检索关键词数量规整到 [1, _MAX_KEYWORD_COUNT]，非法值回退默认 6。"""
@@ -178,6 +184,10 @@ async def researcher_node(state: AIWritingState, config: RunnableConfig) -> dict
 
     def _emit_skill_step(step: dict) -> None:
         skill_steps.append(step)
+        # 技能检索时机 trace：grep "[skill-timing]" 看每步 emit 的真实时刻，
+        # 与前端 [ai-writing:custom] 日志对齐，钉死是否实时推送。
+        logger.info("[skill-timing] t=%.3f emit skill_search_step type=%s skill=%s",
+                    time.monotonic(), step.get("type"), step.get("skill_name") or "")
         writer({"type": "skill_search_step", "agent_name": "素材收集专家", **step})
 
     if is_skill_search:
@@ -205,15 +215,25 @@ async def researcher_node(state: AIWritingState, config: RunnableConfig) -> dict
 
         for entry in selected:
             sk = entry["skill"]
-            # 用 LLM 把 user_query + skill.description 转成针对性关键词
+            skill_query = f"{user_query} {sk.description or ''}".strip() or intent
+            # invoke_skill 步骤带上实际检索词（截断防过长），让用户在每次长 await 之前
+            # 就看到"这一步在用技能 X 检索什么"，配合前端 spinner 体现进行中。
             _emit_skill_step({
                 "type": "invoke_skill",
-                "message": f"使用技能「{sk.name}」开始检索…",
+                "message": f"使用技能「{sk.name}」检索：{skill_query[:60]}",
                 "skill_name": sk.name,
             })
-            skill_query = f"{user_query} {sk.description or ''}".strip() or intent
             try:
-                res = await web_provider.search(skill_query, max_results=per_skill)
+                # 加超时兜底：单个技能检索最多等 _SKILL_SEARCH_TIMEOUT_SECONDS 秒，
+                # 超时不再阻塞后续技能，避免整段过程看起来"冻结"。
+                res = await asyncio.wait_for(
+                    web_provider.search(skill_query, max_results=per_skill),
+                    timeout=_SKILL_SEARCH_TIMEOUT_SECONDS,
+                )
+            except asyncio.TimeoutError:
+                logger.warning("skill[%s] search timed out after %ss", sk.name, _SKILL_SEARCH_TIMEOUT_SECONDS)
+                search_errors.append(f"skill[{sk.name}]: 检索超时（>{_SKILL_SEARCH_TIMEOUT_SECONDS}s）")
+                res = []
             except Exception as e:
                 logger.warning("skill[%s] search failed: %s", sk.name, e)
                 search_errors.append(f"skill[{sk.name}]: {e}")
diff --git a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/search/skill_search.py b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/search/skill_search.py
index d28e79c7..a7be954b 100644
--- a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/search/skill_search.py
+++ b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/search/skill_search.py
@@ -16,6 +16,7 @@
 from __future__ import annotations
 
 import logging
+import time
 from typing import Callable, List, Optional
 
 from langchain_core.messages import HumanMessage, SystemMessage
@@ -140,7 +141,14 @@ class SkillPicker:
         parser = ThinkTagParser()
         full_text = ""
 
+        _chunk_count = {"n": 0}
+
         def _emit_chunk(text: str, is_thinking: bool) -> None:
+            # 只在首个 chunk 打一条 timing，避免逐 token 刷屏；用来和 skill_search_step
+            # 的 [skill-timing] 对比"选择打字"与"步骤事件"谁先到。
+            if _chunk_count["n"] == 0:
+                logger.info("[skill-timing] t=%.3f first skill_select_chunk", time.monotonic())
+            _chunk_count["n"] += 1
             writer({
                 "type": "skill_select_chunk",
                 "text": text,