From d8b43e7579fa6eee597cfcd49345203d7c924c8c Mon Sep 17 00:00:00 2001 From: fengxiangdebaozi <10845145+fengxiangdebaozi@user.noreply.gitee.com> Date: Wed, 24 Jun 2026 22:54:42 +0800 Subject: [PATCH] =?UTF-8?q?fix(ai=E5=86=99=E4=BD=9C):=20=E6=8A=80=E8=83=BD?= =?UTF-8?q?=E6=A3=80=E7=B4=A2=E8=B6=85=E6=97=B6=E7=A1=AC=E7=BC=96=E7=A0=81?= =?UTF-8?q?=205=20=E5=88=86=E9=92=9F=EF=BC=88=E5=86=85=E7=BD=91=E6=85=A2?= =?UTF-8?q?=E6=A3=80=E7=B4=A2=E4=B8=8D=E8=A2=AB=E8=BF=87=E6=97=A9=E5=88=87?= =?UTF-8?q?=E6=96=AD=EF=BC=89?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 内网检索很慢，原 120s 在三个真正生效的层把检索切断。统一硬编码 300s， 不依赖环境变量（部署方不一定会配）： - researcher._SKILL_SEARCH_TIMEOUT_SECONDS：120(env) → 300(写死)，删掉 _env_int； 作用于直连/定向技能检索的 wait_for，也传给真实智能体路径。 - IntranetSearchProvider 默认 httpx 超时：120 → 300（直连内网 ES 的真正瓶颈）。 - run_skill_via_real_agent 整段运行上限：max(timeout_s, 300)，至少 5 分钟。 Co-Authored-By: Claude Opus 4.8 (1M context) --- .../agents/ai_writing/nodes/_skill_agent.py | 9 ++++++++- .../agents/ai_writing/nodes/researcher.py | 16 +++------------- .../agents/ai_writing/search/intranet_stub.py | 4 ++-- 3 files changed, 13 insertions(+), 16 deletions(-) diff --git a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/_skill_agent.py b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/_skill_agent.py index 524dea79..b89c9acd 100644 --- a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/_skill_agent.py +++ b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/_skill_agent.py @@ -52,6 +52,11 @@ def use_skill_search_agent() -> bool: # 单次工具调用循环上限（防跑飞）。 _REAL_AGENT_MAX_TURNS = 16 +# 真实智能体跑技能：整段运行（读 SKILL.md + 多轮工具调用）的超时上限，**硬编码 5 分钟**。 +# 内网检索很慢、技能常做多次检索，单次直连检索的 120s 不够整段用；这里写死，不依赖 env +# （部署方不一定会配环境变量），保证内网慢检索至少有 5 分钟才被切断。 +_REAL_AGENT_TIMEOUT_SECONDS = 300 + _MATERIAL_OUTPUT_CONTRACT = ( "你的任务：严格按上面挂载的检索技能（SKILL.md）的说明，用相应工具检索与「写作意图」相关的**真实素材**。\n" "- 必须按技能说明实际调用工具完成检索（读取知识库文件 / 调用内网检索接口 / MCP 等），**不得凭记忆编造**。\n" @@ -128,7 +133,9 @@ async def run_skill_via_real_agent( skills=[skill_name], # 仅加载这个技能的 SKILL.md model=model_name or "inherit", max_turns=_REAL_AGENT_MAX_TURNS, - timeout_seconds=max(30, timeout_s), + # 整段运行硬编码至少 5 分钟（_REAL_AGENT_TIMEOUT_SECONDS），内网慢检索不被过早切断； + # 调用方传入更大的 timeout_s 时以更大者为准。 + timeout_seconds=max(timeout_s, _REAL_AGENT_TIMEOUT_SECONDS), ) thread_id = (config.get("configurable") or {}).get("thread_id") 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 786b0567..59797efa 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 @@ -86,19 +86,9 @@ _MAX_KEYWORD_COUNT = 10 # 单个技能的检索超时（秒）。技能检索是顺序循环，provider 偶发限速/挂起会把整段 # "技能检索过程"拖住；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。 -# 默认 120s（≥2 分钟）：内网检索 provider 本身就允许 timeout=120，之前 25s 会过早放弃 -# 慢但有效的检索。可用环境变量 AI_WRITING_SKILL_SEARCH_TIMEOUT_SECONDS 覆盖。 -def _env_int(name: str, default: int) -> int: - raw = (os.getenv(name) or "").strip() - if raw: - try: - return max(1, int(raw)) - except ValueError: - pass - return default - - -_SKILL_SEARCH_TIMEOUT_SECONDS = _env_int("AI_WRITING_SKILL_SEARCH_TIMEOUT_SECONDS", 120) +# **硬编码 300s（5 分钟）**：内网检索很慢，给足时间不被过早切断；不依赖环境变量 +# （部署方不一定会配），需要更长时改这里即可。 +_SKILL_SEARCH_TIMEOUT_SECONDS = 300 # 内置「知识库检索」技能名：素材收集专家 agent 的默认技能。researcher 对它 # 特殊处理——不用 description 拼 query，而是对 Step 1 生成的全部检索词做 diff --git a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/search/intranet_stub.py b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/search/intranet_stub.py index 7a8bf1e3..00f7f042 100644 --- a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/search/intranet_stub.py +++ b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/search/intranet_stub.py @@ -27,12 +27,12 @@ class IntranetSearchProvider(BaseSearchProvider): self, base_url: str = "", knowledge_base_name: str = _DEFAULT_KB, - timeout: int = 120, + timeout: int = 300, **_ignored, ): self.endpoint = base_url or _DEFAULT_ENDPOINT self.knowledge_base_name = knowledge_base_name or _DEFAULT_KB - # 内网 ES 接口可能较慢，超时放宽到 2 分钟 + # 内网 ES 接口很慢，超时硬编码放宽到 5 分钟（不依赖 env 配置） self.timeout = timeout # 收集本次运行的检索异常，供 researcher 节点取出并提示用户（ES 波动等） self.errors: list[str] = []