feat(pool,scheduling): 号池调度维度、配额冷却机制与管理后台重构

- 新增 scheduling_dimensions 模块，为每个 Key 计算多维调度状态(手动/冷却/熔断/成本/健康)
- 新增 quota_cooldown 模块，统一判定 Key 的有效冷却原因
- Pool 管理后台 API 扩展 Key 详情字段(调度状态/维度/配额/OAuth 信息)
- 前端 Pool 管理页面重写，支持调度状态展示、批量清理封禁 Key
- Handler 基类增加请求调度元数据采集，stream telemetry 增强
- 请求时间线组件增强，支持 attempted 候选展示
- Kiro OAuth 凭证导入解析改进
- 新增 usage 表 provider_key 索引迁移
- 补充调度维度、配额冷却、候选枚举等单元测试

Closes #197

Co-authored-by: AAEE86 <33052466+AAEE86@users.noreply.github.com>

tests/unit/test_request_scheduling_metadata.py

@@ -0,0 +1,154 @@
+from src.api.handlers.base.base_handler import BaseMessageHandler
+from src.services.candidate.schema import CandidateKey
+
+
+def _handler() -> BaseMessageHandler:
+    return BaseMessageHandler.__new__(BaseMessageHandler)
+
+
+def test_scheduling_audit_detects_internal_failover() -> None:
+    handler = _handler()
+    metadata = handler._merge_scheduling_metadata(
+        {},
+        candidate_keys=[
+            {
+                "candidate_index": 0,
+                "retry_index": 0,
+                "provider_id": "p1",
+                "provider_name": "provider-a",
+                "key_id": "k1",
+                "key_name": "account-a",
+                "status": "failed",
+                "status_code": 429,
+            },
+            {
+                "candidate_index": 1,
+                "retry_index": 0,
+                "provider_id": "p1",
+                "provider_name": "provider-a",
+                "key_id": "k2",
+                "key_name": "account-b",
+                "status": "success",
+                "status_code": 200,
+            },
+        ],
+        selected_key_id="k2",
+        fallback_from_request=False,
+    )
+    assert metadata is not None
+    audit = metadata.get("scheduling_audit")
+    assert isinstance(audit, dict)
+    assert audit.get("attempted_count") == 2
+    assert audit.get("account_count") == 2
+    assert audit.get("retry_occurred") is True
+    assert audit.get("failover_occurred") is True
+    assert audit.get("selected_key_id") == "k2"
+
+    accounts = audit.get("accounts")
+    assert isinstance(accounts, list)
+    assert any(a.get("key_id") == "k2" and a.get("selected") for a in accounts)
+
+
+def test_scheduling_audit_distinguishes_retry_from_failover() -> None:
+    handler = _handler()
+    metadata = handler._merge_scheduling_metadata(
+        {},
+        candidate_keys=[
+            {
+                "candidate_index": 0,
+                "retry_index": 0,
+                "provider_id": "p1",
+                "provider_name": "provider-a",
+                "key_id": "k1",
+                "key_name": "account-a",
+                "status": "failed",
+                "status_code": 503,
+            },
+            {
+                "candidate_index": 0,
+                "retry_index": 1,
+                "provider_id": "p1",
+                "provider_name": "provider-a",
+                "key_id": "k1",
+                "key_name": "account-a",
+                "status": "success",
+                "status_code": 200,
+            },
+        ],
+        selected_key_id="k1",
+        fallback_from_request=False,
+    )
+    assert metadata is not None
+    audit = metadata.get("scheduling_audit")
+    assert isinstance(audit, dict)
+    assert audit.get("attempted_count") == 2
+    assert audit.get("account_count") == 1
+    assert audit.get("retry_occurred") is True
+    assert audit.get("failover_occurred") is False
+
+
+def test_scheduling_metadata_supports_candidate_key_dataclass() -> None:
+    handler = _handler()
+    metadata = handler._merge_scheduling_metadata(
+        {},
+        candidate_keys=[
+            CandidateKey(
+                candidate_index=0,
+                retry_index=0,
+                provider_id="p1",
+                provider_name="provider-a",
+                endpoint_id="e1",
+                key_id="k1",
+                key_name="account-a",
+                status="success",
+                status_code=200,
+            )
+        ],
+        selected_key_id="k1",
+        fallback_from_request=False,
+    )
+    assert metadata is not None
+    snapshots = metadata.get("candidate_keys")
+    assert isinstance(snapshots, list)
+    assert snapshots[0]["status"] == "success"
+    assert snapshots[0]["key_id"] == "k1"
+
+
+def test_scheduling_audit_excludes_unused_candidates() -> None:
+    handler = _handler()
+    metadata = handler._merge_scheduling_metadata(
+        {},
+        candidate_keys=[
+            {
+                "candidate_index": 0,
+                "retry_index": 0,
+                "provider_id": "p1",
+                "provider_name": "provider-a",
+                "key_id": "k1",
+                "key_name": "account-a",
+                "status": "success",
+                "status_code": 200,
+            },
+            {
+                "candidate_index": 1,
+                "retry_index": 0,
+                "provider_id": "p1",
+                "provider_name": "provider-a",
+                "key_id": "k2",
+                "key_name": "account-b",
+                "status": "unused",
+            },
+        ],
+        selected_key_id="k1",
+        fallback_from_request=False,
+    )
+
+    assert metadata is not None
+    audit = metadata.get("scheduling_audit")
+    assert isinstance(audit, dict)
+    assert audit.get("attempted_count") == 1
+    assert audit.get("account_count") == 1
+    attempts = audit.get("attempts")
+    assert isinstance(attempts, list)
+    assert len(attempts) == 1
+    assert attempts[0].get("key_id") == "k1"