feat(ops): 实现上游错误事件记录与查询功能

**新增功能**:
- 新建ops_upstream_error_events表存储上游服务错误详情
- 支持记录上游429/529/5xx错误的详细上下文信息
- 提供按时间范围查询上游错误事件的API

**后端改动**:
1. 模型层（ops_models.go, ops_port.go）:
   - 新增UpstreamErrorEvent结构体
   - 扩展Repository接口支持上游错误事件CRUD

2. 仓储层（ops_repo.go）:
   - 实现InsertUpstreamErrorEvent写入上游错误
   - 实现GetUpstreamErrorEvents按时间范围查询

3. 服务层（ops_service.go, ops_upstream_context.go）:
   - ops_service: 新增GetUpstreamErrorEvents查询方法
   - ops_upstream_context: 封装上游错误上下文构建逻辑

4. Handler层（ops_error_logger.go）:
   - 新增GetUpstreamErrorsHandler处理上游错误查询请求

5. Gateway层集成:
   - antigravity_gateway_service.go: 429/529错误时记录上游事件
   - gateway_service.go: OpenAI 429/5xx错误时记录
   - gemini_messages_compat_service.go: Gemini 429/5xx错误时记录
   - openai_gateway_service.go: OpenAI 429/5xx错误时记录
   - ratelimit_service.go: 429限流错误时记录

**数据记录字段**:
- request_id: 关联ops_logs主记录
- platform/model: 上游服务标识
- status_code/error_message: 错误详情
- request_headers/response_body: 调试信息（可选）
- created_at: 错误发生时间

backend/internal/service/ops_service.go

@@ -163,6 +163,61 @@ func (s *OpsService) RecordError(ctx context.Context, entry *OpsInsertErrorLogIn
 		}
 	}
 
+	// Sanitize + serialize upstream error events list.
+	if len(entry.UpstreamErrors) > 0 {
+		const maxEvents = 32
+		events := entry.UpstreamErrors
+		if len(events) > maxEvents {
+			events = events[len(events)-maxEvents:]
+		}
+
+		sanitized := make([]*OpsUpstreamErrorEvent, 0, len(events))
+		for _, ev := range events {
+			if ev == nil {
+				continue
+			}
+			out := *ev
+
+			out.Platform = strings.TrimSpace(out.Platform)
+			out.UpstreamRequestID = truncateString(strings.TrimSpace(out.UpstreamRequestID), 128)
+			out.Kind = truncateString(strings.TrimSpace(out.Kind), 64)
+
+			if out.AccountID < 0 {
+				out.AccountID = 0
+			}
+			if out.UpstreamStatusCode < 0 {
+				out.UpstreamStatusCode = 0
+			}
+			if out.AtUnixMs < 0 {
+				out.AtUnixMs = 0
+			}
+
+			msg := sanitizeUpstreamErrorMessage(strings.TrimSpace(out.Message))
+			msg = truncateString(msg, 2048)
+			out.Message = msg
+
+			detail := strings.TrimSpace(out.Detail)
+			if detail != "" {
+				// Keep upstream detail small; request bodies are not stored here, only upstream error payloads.
+				sanitizedDetail, _ := sanitizeErrorBodyForStorage(detail, opsMaxStoredErrorBodyBytes)
+				out.Detail = sanitizedDetail
+			} else {
+				out.Detail = ""
+			}
+
+			// Drop fully-empty events (can happen if only status code was known).
+			if out.UpstreamStatusCode == 0 && out.Message == "" && out.Detail == "" {
+				continue
+			}
+
+			evCopy := out
+			sanitized = append(sanitized, &evCopy)
+		}
+
+		entry.UpstreamErrorsJSON = marshalOpsUpstreamErrors(sanitized)
+		entry.UpstreamErrors = nil
+	}
+
 	if _, err := s.opsRepo.InsertErrorLog(ctx, entry); err != nil {
 		// Never bubble up to gateway; best-effort logging.
 		log.Printf("[Ops] RecordError failed: %v", err)