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/tests/unit/test_phase5_skills_runtime.py

@@ -298,8 +298,29 @@ def test_skill_learning_service_generates_candidates_and_retire_draft(tmp_path:
                 ended_at=recent,
                 success=True,
                 finish_reason="stop",
+                feedback={"feedback_type": "satisfied"},
+                activated_skills=[],
+                task_id=f"task-new-{index}",
+                attempt_index=1,
+                validation_result={"accepted": True, "score": 0.9},
+            )
+        )
+
+    for index in range(2):
+        run_store.append_run_record(
+            RunRecord(
+                run_id=f"simple-chat-{index}",
+                session_id="session-simple",
+                task_text="你是谁",
+                started_at=recent,
+                ended_at=recent,
+                success=True,
+                finish_reason="stop",
                 feedback={},
                 activated_skills=[],
+                task_id=None,
+                attempt_index=None,
+                validation_result=None,
             )
         )
 
@@ -329,8 +350,11 @@ def test_skill_learning_service_generates_candidates_and_retire_draft(tmp_path:
                 ended_at=recent,
                 success=True,
                 finish_reason="stop",
-                feedback={},
+                feedback={"feedback_type": "satisfied"},
                 activated_skills=receipts,
+                task_id=f"task-merge-{index}",
+                attempt_index=1,
+                validation_result={"accepted": True, "score": 0.9},
             )
         )
         for receipt in receipts:
@@ -382,6 +406,9 @@ def test_skill_learning_service_generates_candidates_and_retire_draft(tmp_path:
     kinds = {candidate.kind for candidate in candidates}
 
     assert {"revise_skill", "new_skill", "merge_skills", "retire_skill"} <= kinds
+    new_candidates = [candidate for candidate in candidates if candidate.kind == "new_skill"]
+    assert new_candidates
+    assert all("simple-chat" not in run_id for candidate in new_candidates for run_id in candidate.source_run_ids)
 
     retire_candidate = next(candidate for candidate in candidates if candidate.kind == "retire_skill")
     retire_draft = asyncio.run(
@@ -396,6 +423,100 @@ def test_skill_learning_service_generates_candidates_and_retire_draft(tmp_path:
     assert store.read_draft("svn-migration", retire_draft.draft_id) is not None
 
 
+def test_skill_learning_service_generates_task_scoped_candidates(tmp_path: Path) -> None:
+    store = SkillSpecStore(tmp_path)
+    run_store = RunMemoryStore(tmp_path / "memory" / "runs")
+    learning_store = SkillLearningStore(tmp_path / "memory" / "skills")
+    service = SkillLearningService(
+        run_store=run_store,
+        learning_store=learning_store,
+        draft_service=DraftService(store),
+        evidence_selector=EvidenceSelector(run_store),
+    )
+    now = datetime.now(timezone.utc).isoformat()
+    receipt = _receipt(
+        run_id="task-run-1",
+        session_id="session-task",
+        skill_name="api-review",
+        skill_version="v0001",
+        activated_at=now,
+    )
+    run_store.append_run_record(
+        RunRecord(
+            run_id="task-run-1",
+            session_id="session-task",
+            task_id="task-1",
+            attempt_index=1,
+            task_text="Review API compatibility",
+            started_at=now,
+            ended_at=now,
+            success=True,
+            finish_reason="stop",
+            feedback={"feedback_type": "satisfied"},
+            activated_skills=[receipt],
+            validation_result={"accepted": True, "score": 0.9},
+        )
+    )
+    run_store.append_run_record(
+        RunRecord(
+            run_id="other-task-run",
+            session_id="session-other",
+            task_id="task-2",
+            attempt_index=1,
+            task_text="Review API compatibility",
+            started_at=now,
+            ended_at=now,
+            success=True,
+            finish_reason="stop",
+            feedback={"feedback_type": "satisfied"},
+            activated_skills=[],
+            validation_result={"accepted": True, "score": 0.9},
+        )
+    )
+
+    candidates = service.build_learning_candidates_for_task("task-1", trigger_run_id="task-run-1")
+
+    assert [candidate.candidate_id for candidate in candidates] == ["revise:api-review:v0001:task:task-1"]
+    assert candidates[0].source_run_ids == ["task-run-1"]
+    assert candidates[0].related_skill_names == ["api-review"]
+    assert candidates[0].evidence["task_id"] == "task-1"
+
+
+def test_skill_learning_service_generates_new_skill_for_task_without_published_skills(tmp_path: Path) -> None:
+    store = SkillSpecStore(tmp_path)
+    run_store = RunMemoryStore(tmp_path / "memory" / "runs")
+    learning_store = SkillLearningStore(tmp_path / "memory" / "skills")
+    service = SkillLearningService(
+        run_store=run_store,
+        learning_store=learning_store,
+        draft_service=DraftService(store),
+        evidence_selector=EvidenceSelector(run_store),
+    )
+    now = datetime.now(timezone.utc).isoformat()
+    run_store.append_run_record(
+        RunRecord(
+            run_id="task-run-1",
+            session_id="session-task",
+            task_id="task-1",
+            attempt_index=1,
+            task_text="Generate migration checklist",
+            started_at=now,
+            ended_at=now,
+            success=True,
+            finish_reason="stop",
+            feedback={"feedback_type": "satisfied"},
+            activated_skills=[],
+            validation_result={"accepted": True, "score": 0.9},
+        )
+    )
+
+    candidates = service.build_learning_candidates_for_task("task-1", trigger_run_id="task-run-1")
+
+    assert [candidate.candidate_id for candidate in candidates] == ["new:task:task-1"]
+    assert candidates[0].kind == "new_skill"
+    assert candidates[0].source_run_ids == ["task-run-1"]
+
+
 def test_agent_loop_records_skill_receipts_and_effects(tmp_path: Path) -> None:
     skill = SkillContext(
         name="docker-debug",
@@ -446,7 +567,7 @@ def test_agent_loop_records_skill_receipts_and_effects(tmp_path: Path) -> None:
     skill_effects = next(event for event in events if event.event_type == "skill_effects_snapshotted")
     assert skill_effects.event_payload["run_record"]["activated_skills"][0]["skill_version"] == "v0007"
     assert skill_effects.event_payload["skill_effects"][0]["skill_name"] == "docker-debug"
-    assert skill_effects.event_payload["learning_candidate_enabled"] is False
+    assert skill_effects.event_payload["candidate_generation_allowed"] is False
     assert skill_effects.event_payload["learning_candidates"] == []
 
     run_records = loaded.run_memory_store.list_runs()