feat(engine): 添加MCP连接管理和工具集成功能

- 集成MCP连接管理器，支持MCP服务器连接
- 添加多种内置工具：ClarifyTool、CronTool、DelegateTool、ExecuteCodeTool、
  PatchFileTool、ProcessTool、SendMessageTool、SpawnTool、TerminalTool、
  TodoTool、WebFetchTool、WebSearchTool、WriteFileTool等
- 实现工具注册和装配功能
- 添加技能选择上下文参数
- 支持思考模式控制参数thinking_enabled

feat(coordinator): 重构任务执行计划器参数命名

- 将learning_candidate_enabled重命名为allow_candidate_generation
- 更新TeamGraphScheduler中的参数传递
- 修改LocalAgentRunner中的相关参数处理
- 更新README文档中的相应描述

refactor(context): 标准化工具调用参数格式

- 添加_json导入用于参数序列化
- 实现_provider_tool_calls方法标准化OpenAI兼容的工具调用载荷
- 修复工具调用中参数非字符串类型的序列化问题

refactor(session): 优化消息历史记录过滤逻辑

- 修改get_messages_as_conversation为基于运行状态过滤消息
- 排除未完成、失败或错误结束的运行记录
- 改进对话历史的可见性控制机制

fix(store): 修复FTS索引重建逻辑

- 添加异常处理防止FTS索引创建失败
- 实现_rebuild_fts_index方法重新构建全文搜索索引
- 优化索引触发器和表的维护流程

app-instance/backend/beaver/skills/learning/service.py

@@ -69,6 +69,94 @@ class SkillLearningService:
                 existing_ids.add(candidate.candidate_id)
         return candidates
 
+    def build_learning_candidates_for_task(self, task_id: str, *, trigger_run_id: str) -> list[SkillLearningCandidate]:
+        """Build candidates scoped to a single validated and satisfied Task run."""
+
+        runs = [record for record in self.run_store.list_runs() if record.task_id == task_id]
+        trigger_run = next((record for record in runs if record.run_id == trigger_run_id), None)
+        if trigger_run is None or not self._is_confirmed_positive_run(trigger_run):
+            return []
+
+        source_runs = [record for record in runs if self._is_confirmed_positive_run(record)]
+        if not source_runs:
+            return []
+
+        candidates: list[SkillLearningCandidate] = []
+        published_receipts = [
+            receipt
+            for record in source_runs
+            for receipt in record.activated_skills
+            if self._is_published_skill_receipt(receipt)
+        ]
+        source_run_ids = [record.run_id for record in source_runs]
+        source_session_ids = list(dict.fromkeys(record.session_id for record in source_runs))
+
+        if not published_receipts:
+            candidates.append(
+                SkillLearningCandidate(
+                    candidate_id=f"new:task:{task_id}",
+                    kind="new_skill",
+                    source_run_ids=source_run_ids,
+                    source_session_ids=source_session_ids,
+                    related_skill_names=[],
+                    reason=f"Task {task_id} completed successfully without a published skill; consider extracting reusable guidance.",
+                    evidence={"task_id": task_id, "trigger_run_id": trigger_run_id, "theme": self._task_theme(trigger_run.task_text)},
+                    status="open",
+                    priority=1,
+                    confidence=0.8,
+                    trigger_reason="validation_accepted_and_user_satisfied",
+                )
+            )
+        else:
+            seen: set[tuple[str, str]] = set()
+            for receipt in published_receipts:
+                key = (receipt.skill_name, receipt.skill_version)
+                if key in seen:
+                    continue
+                seen.add(key)
+                skill_runs = [
+                    record
+                    for record in source_runs
+                    if any(
+                        item.skill_name == receipt.skill_name
+                        and item.skill_version == receipt.skill_version
+                        and self._is_published_skill_receipt(item)
+                        for item in record.activated_skills
+                    )
+                ]
+                candidates.append(
+                    SkillLearningCandidate(
+                        candidate_id=f"revise:{receipt.skill_name}:{receipt.skill_version}:task:{task_id}",
+                        kind="revise_skill",
+                        source_run_ids=[record.run_id for record in skill_runs],
+                        source_session_ids=list(dict.fromkeys(record.session_id for record in skill_runs)),
+                        related_skill_names=[receipt.skill_name],
+                        reason=(
+                            f"Task {task_id} succeeded with published skill "
+                            f"{receipt.skill_name}/{receipt.skill_version}; consider whether the skill should capture this evidence."
+                        ),
+                        evidence={
+                            "task_id": task_id,
+                            "trigger_run_id": trigger_run_id,
+                            "skill_version": receipt.skill_version,
+                        },
+                        status="open",
+                        priority=1,
+                        confidence=0.7,
+                        trigger_reason="validation_accepted_and_user_satisfied",
+                    )
+                )
+
+        existing_ids = {item.candidate_id for item in self.learning_store.list_learning_candidates()}
+        created: list[SkillLearningCandidate] = []
+        for candidate in candidates:
+            if candidate.candidate_id in existing_ids:
+                continue
+            self.learning_store.record_learning_candidate(candidate)
+            existing_ids.add(candidate.candidate_id)
+            created.append(candidate)
+        return created
+
     async def synthesize_draft(self, candidate_id: str, provider_bundle: ProviderBundle) -> Any:
         candidates = {item.candidate_id: item for item in self.learning_store.list_learning_candidates()}
         candidate = candidates.get(candidate_id)
@@ -181,7 +269,7 @@ class SkillLearningService:
             groups.setdefault(key, []).append(record)
         candidates: list[SkillLearningCandidate] = []
         for theme, runs in groups.items():
-            successful = [record for record in runs if record.success]
+            successful = [record for record in runs if self._is_confirmed_positive_run(record)]
             if len(successful) < 2:
                 continue
             if any(record.activated_skills for record in successful):
@@ -202,6 +290,8 @@ class SkillLearningService:
     def _build_merge_candidates(self) -> list[SkillLearningCandidate]:
         pair_counts: dict[tuple[str, str], list[RunRecord]] = {}
         for record in self.run_store.list_runs():
+            if not self._is_confirmed_positive_run(record):
+                continue
             unique = sorted({receipt.skill_name for receipt in record.activated_skills})
             for pair in combinations(unique, 2):
                 pair_counts.setdefault(pair, []).append(record)
@@ -260,6 +350,25 @@ class SkillLearningService:
                 effects.extend(self.run_store.list_skill_effects(receipt.skill_name, version=receipt.skill_version))
         return effects
 
+    @staticmethod
+    def _is_confirmed_positive_run(record: RunRecord) -> bool:
+        validation = record.validation_result or {}
+        feedback = record.feedback or {}
+        return (
+            bool(record.success)
+            and bool(record.task_id)
+            and validation.get("accepted") is True
+            and feedback.get("feedback_type") == "satisfied"
+        )
+
+    @staticmethod
+    def _is_published_skill_receipt(receipt: SkillActivationReceipt) -> bool:
+        return (
+            not receipt.skill_name.startswith(("draft:", "ephemeral:"))
+            and not receipt.skill_version.startswith(("draft:", "ephemeral:"))
+            and receipt.activation_reason not in {"generated_missing_skill", "ephemeral_guidance"}
+        )
+
     @staticmethod
     def _candidate_id(kind: str, *parts: str) -> str:
         return f"{kind}:{'|'.join(parts)}"