feat(task): 添加任务修订功能和超时处理机制

添加了 `revise_task` 路由动作类型，允许用户修改、纠正或重新执行最新活动任务结果。
实现了工具失败指导原则，防止相同类别工具重复失败。
为任务规划器添加了超时处理机制，避免长时间等待。

BREAKING CHANGE: 任务路由逻辑已更新，新增 `revise_task` 动作类型。

fix(api): 修复任务详情API返回完整流程投影

修复了任务详情API端点，现在会包含过滤后的流程运行、事件和工件信息，
并确保时间戳字段正确序列化。

refactor(engine): 优化任务技能解析器摘要节点处理

改进了任务技能解析器对摘要节点的处理逻辑，对于仅依赖文本生成功能的摘要节
点不再分配具体技能，直接使用依赖项输出进行汇总。

test: 增加任务修订和超时处理测试用例

添加了测试用例验证任务修订输入记录反馈、超时回退到单模式以及
摘要节点技能解析等新功能。

app-instance/backend/beaver/tasks/skill_resolver.py

@@ -93,6 +93,29 @@ class TaskSkillResolver:
             for item in node.agent.metadata.get("required_capabilities", [])
             if str(item).strip()
         ]
+        if self._is_summary_only_node(node, skill_query=skill_query, required_capabilities=required_capabilities):
+            resolved = self._generic_node(
+                node,
+                pinned_skill_names=[],
+                pinned_skill_contexts=[],
+                metadata={
+                    **node.agent.metadata,
+                    "skill_query": skill_query,
+                    "required_capabilities": required_capabilities,
+                    "selected_skill_names": [],
+                    "ephemeral_skill_names": [],
+                    "summary_uses_dependency_outputs_only": True,
+                },
+            )
+            return resolved, SkillResolutionReport(
+                node_id=node.node_id,
+                skill_query=skill_query,
+                required_capabilities=required_capabilities,
+                selected_skill_names=[],
+                ephemeral_used=False,
+                reason="summary node uses dependency outputs directly",
+            )
+
         selected = await self._select_published_skills(
             query="\n".join(
                 part
@@ -226,6 +249,34 @@ class TaskSkillResolver:
                 selected.append(name)
         return selected
 
+    @staticmethod
+    def _is_summary_only_node(
+        node: ExecutionNode,
+        *,
+        skill_query: str,
+        required_capabilities: list[str],
+    ) -> bool:
+        node_id = node.node_id.strip().lower()
+        query = skill_query.strip().lower()
+        capabilities = {item.strip().lower() for item in required_capabilities}
+        task_text = node.task.strip().lower()
+        summary_identity = node_id in {"summarize", "summary", "synthesis"} or query in {
+            "summarization",
+            "summary",
+            "synthesis",
+            "final synthesis",
+        }
+        text_only_capabilities = not capabilities or capabilities.issubset(
+            {"text generation", "summarization", "summary", "synthesis"}
+        )
+        dependency_summary_task = (
+            "summary" in task_text
+            or "summarize" in task_text
+            or "synthesis" in task_text
+            or "compile" in task_text
+        )
+        return summary_identity and text_only_capabilities and dependency_summary_task
+
     @staticmethod
     def _generic_node(
         node: ExecutionNode,
@@ -246,7 +297,9 @@ class TaskSkillResolver:
                 },
             ),
             inherited_pinned_skills=pinned_skill_names,
-            inherited_pinned_skill_contexts=list(pinned_skill_contexts or node.inherited_pinned_skill_contexts),
+            inherited_pinned_skill_contexts=list(
+                node.inherited_pinned_skill_contexts if pinned_skill_contexts is None else pinned_skill_contexts
+            ),
         )
 
     @staticmethod