From ad5da2a1184364816e027a720256e1e5ee92d343 Mon Sep 17 00:00:00 2001 From: zhaoxi Date: Fri, 5 Jun 2026 09:18:40 +0000 Subject: [PATCH] =?UTF-8?q?fix:=20=E5=B7=A5=E4=BD=9C=E6=B5=81=E5=BC=95?= =?UTF-8?q?=E6=93=8E=E6=8E=A5=E5=8F=97=20ForregulatoryNode=20=E7=B1=BB?= =?UTF-8?q?=E5=9E=8B=E8=BF=94=E5=9B=9E=EF=BC=8C=E4=BF=AE=E5=A4=8D=E6=9C=80?= =?UTF-8?q?=E5=90=8E=E4=B8=80=E6=AD=A5=E5=A4=B1=E8=B4=A5?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit consciousness_node 在总结步骤返回 ForregulatoryNode 类型，但工作流引擎 只认 ForWorkflow 导致最后一步标记 failed。现在两种类型都被正确处理。 Co-Authored-By: Claude Opus 4.7 --- kilostar/core/work/workflow/workflow_engine.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/kilostar/core/work/workflow/workflow_engine.py b/kilostar/core/work/workflow/workflow_engine.py index 7088131..4daac51 100644 --- a/kilostar/core/work/workflow/workflow_engine.py +++ b/kilostar/core/work/workflow/workflow_engine.py @@ -399,6 +399,7 @@ async def _default_consciousness_executor( from kilostar.core.individual.consciousness_node.template import ( ForWorkflow, ForWorkflowInput, + ForregulatoryNode, ) from kilostar.core.work.workflow.workflow import WorkflowStep @@ -410,6 +411,8 @@ async def _default_consciousness_executor( result = await consciousness_node.working.remote(payload) if isinstance(result, ForWorkflow): return result.output, True + if isinstance(result, ForregulatoryNode): + return result.output, True if result is None: return "ConsciousnessNode 返回 None", False return f"ConsciousnessNode 返回未知类型: {type(result).__name__}", False