Merge branch 'Wei-Shaw:main' into fix/open-issues-cleanup

backend/internal/service/account_usage_service.go

@@ -45,6 +45,8 @@ type UsageLogRepository interface {
 	GetDashboardStats(ctx context.Context) (*usagestats.DashboardStats, error)
 	GetUsageTrendWithFilters(ctx context.Context, startTime, endTime time.Time, granularity string, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) ([]usagestats.TrendDataPoint, error)
 	GetModelStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, requestType *int16, stream *bool, billingType *int8) ([]usagestats.ModelStat, error)
+	GetEndpointStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) ([]usagestats.EndpointStat, error)
+	GetUpstreamEndpointStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) ([]usagestats.EndpointStat, error)
 	GetGroupStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, requestType *int16, stream *bool, billingType *int8) ([]usagestats.GroupStat, error)
 	GetAPIKeyUsageTrend(ctx context.Context, startTime, endTime time.Time, granularity string, limit int) ([]usagestats.APIKeyUsageTrendPoint, error)
 	GetUserUsageTrend(ctx context.Context, startTime, endTime time.Time, granularity string, limit int) ([]usagestats.UserUsageTrendPoint, error)

backend/internal/service/admin_service.go

@@ -832,7 +832,7 @@ func (s *adminServiceImpl) CreateGroup(ctx context.Context, input *CreateGroupIn
 		subscriptionType = SubscriptionTypeStandard
 	}
 
-	// 限额字段：0 和 nil 都表示"无限制"
+	// 限额字段：nil/负数 表示"无限制"，0 表示"不允许用量"，正数表示具体限额
 	dailyLimit := normalizeLimit(input.DailyLimitUSD)
 	weeklyLimit := normalizeLimit(input.WeeklyLimitUSD)
 	monthlyLimit := normalizeLimit(input.MonthlyLimitUSD)
@@ -944,9 +944,9 @@ func (s *adminServiceImpl) CreateGroup(ctx context.Context, input *CreateGroupIn
 	return group, nil
 }
 
-// normalizeLimit 将 0 或负数转换为 nil（表示无限制）
+// normalizeLimit 将负数转换为 nil（表示无限制），0 保留（表示限额为零）
 func normalizeLimit(limit *float64) *float64 {
-	if limit == nil || *limit <= 0 {
+	if limit == nil || *limit < 0 {
 		return nil
 	}
 	return limit
@@ -1058,16 +1058,11 @@ func (s *adminServiceImpl) UpdateGroup(ctx context.Context, id int64, input *Upd
 	if input.SubscriptionType != "" {
 		group.SubscriptionType = input.SubscriptionType
 	}
-	// 限额字段：0 和 nil 都表示"无限制"，正数表示具体限额
-	if input.DailyLimitUSD != nil {
-		group.DailyLimitUSD = normalizeLimit(input.DailyLimitUSD)
-	}
-	if input.WeeklyLimitUSD != nil {
-		group.WeeklyLimitUSD = normalizeLimit(input.WeeklyLimitUSD)
-	}
-	if input.MonthlyLimitUSD != nil {
-		group.MonthlyLimitUSD = normalizeLimit(input.MonthlyLimitUSD)
-	}
+	// 限额字段：nil/负数 表示"无限制"，0 表示"不允许用量"，正数表示具体限额
+	// 前端始终发送这三个字段，无需 nil 守卫
+	group.DailyLimitUSD = normalizeLimit(input.DailyLimitUSD)
+	group.WeeklyLimitUSD = normalizeLimit(input.WeeklyLimitUSD)
+	group.MonthlyLimitUSD = normalizeLimit(input.MonthlyLimitUSD)
 	// 图片生成计费配置：负数表示清除（使用默认价格）
 	if input.ImagePrice1K != nil {
 		group.ImagePrice1K = normalizePrice(input.ImagePrice1K)

backend/internal/service/api_key.go

@@ -22,8 +22,9 @@ const (
 )
 
 // IsWindowExpired returns true if the window starting at windowStart has exceeded the given duration.
+// A nil windowStart is treated as expired — no initialized window means any accumulated usage is stale.
 func IsWindowExpired(windowStart *time.Time, duration time.Duration) bool {
-	return windowStart != nil && time.Since(*windowStart) >= duration
+	return windowStart == nil || time.Since(*windowStart) >= duration
 }
 
 type APIKey struct {

backend/internal/service/api_key_rate_limit_test.go

@@ -15,10 +15,10 @@ func TestIsWindowExpired(t *testing.T) {
 		want     bool
 	}{
 		{
-			name:     "nil window start",
+			name:     "nil window start (treated as expired)",
 			start:    nil,
 			duration: RateLimitWindow5h,
-			want:     false,
+			want:     true,
 		},
 		{
 			name:     "active window (started 1h ago, 5h window)",
@@ -113,7 +113,7 @@ func TestAPIKey_EffectiveUsage(t *testing.T) {
 			want7d: 0,
 		},
 		{
-			name: "nil window starts return raw usage",
+			name: "nil window starts return 0 (stale usage reset)",
 			key: APIKey{
 				Usage5h:       5.0,
 				Usage1d:       10.0,
@@ -122,9 +122,9 @@ func TestAPIKey_EffectiveUsage(t *testing.T) {
 				Window1dStart: nil,
 				Window7dStart: nil,
 			},
-			want5h: 5.0,
-			want1d: 10.0,
-			want7d: 50.0,
+			want5h: 0,
+			want1d: 0,
+			want7d: 0,
 		},
 		{
 			name: "mixed: 5h expired, 1d active, 7d nil",
@@ -138,7 +138,7 @@ func TestAPIKey_EffectiveUsage(t *testing.T) {
 			},
 			want5h: 0,
 			want1d: 10.0,
-			want7d: 50.0,
+			want7d: 0,
 		},
 		{
 			name: "zero usage with active windows",
@@ -210,7 +210,7 @@ func TestAPIKeyRateLimitData_EffectiveUsage(t *testing.T) {
 			want7d: 0,
 		},
 		{
-			name: "nil window starts return raw usage",
+			name: "nil window starts return 0 (stale usage reset)",
 			data: APIKeyRateLimitData{
 				Usage5h:       3.0,
 				Usage1d:       8.0,
@@ -219,9 +219,9 @@ func TestAPIKeyRateLimitData_EffectiveUsage(t *testing.T) {
 				Window1dStart: nil,
 				Window7dStart: nil,
 			},
-			want5h: 3.0,
-			want1d: 8.0,
-			want7d: 40.0,
+			want5h: 0,
+			want1d: 0,
+			want7d: 0,
 		},
 	}
 

backend/internal/service/gateway_record_usage_test.go

@@ -369,3 +369,54 @@ func TestGatewayServiceRecordUsage_BillingErrorSkipsUsageLogWrite(t *testing.T)
 	require.Equal(t, 1, billingRepo.calls)
 	require.Equal(t, 0, usageRepo.calls)
 }
+
+func TestGatewayServiceRecordUsage_ReasoningEffortPersisted(t *testing.T) {
+	usageRepo := &openAIRecordUsageBestEffortLogRepoStub{}
+	svc := newGatewayRecordUsageServiceForTest(usageRepo, &openAIRecordUsageUserRepoStub{}, &openAIRecordUsageSubRepoStub{})
+
+	effort := "max"
+	err := svc.RecordUsage(context.Background(), &RecordUsageInput{
+		Result: &ForwardResult{
+			RequestID: "effort_test",
+			Usage: ClaudeUsage{
+				InputTokens:  10,
+				OutputTokens: 5,
+			},
+			Model:           "claude-opus-4-6",
+			Duration:        time.Second,
+			ReasoningEffort: &effort,
+		},
+		APIKey:  &APIKey{ID: 1},
+		User:    &User{ID: 1},
+		Account: &Account{ID: 1},
+	})
+
+	require.NoError(t, err)
+	require.NotNil(t, usageRepo.lastLog)
+	require.NotNil(t, usageRepo.lastLog.ReasoningEffort)
+	require.Equal(t, "max", *usageRepo.lastLog.ReasoningEffort)
+}
+
+func TestGatewayServiceRecordUsage_ReasoningEffortNil(t *testing.T) {
+	usageRepo := &openAIRecordUsageBestEffortLogRepoStub{}
+	svc := newGatewayRecordUsageServiceForTest(usageRepo, &openAIRecordUsageUserRepoStub{}, &openAIRecordUsageSubRepoStub{})
+
+	err := svc.RecordUsage(context.Background(), &RecordUsageInput{
+		Result: &ForwardResult{
+			RequestID: "no_effort_test",
+			Usage: ClaudeUsage{
+				InputTokens:  10,
+				OutputTokens: 5,
+			},
+			Model:    "claude-sonnet-4",
+			Duration: time.Second,
+		},
+		APIKey:  &APIKey{ID: 1},
+		User:    &User{ID: 1},
+		Account: &Account{ID: 1},
+	})
+
+	require.NoError(t, err)
+	require.NotNil(t, usageRepo.lastLog)
+	require.Nil(t, usageRepo.lastLog.ReasoningEffort)
+}

backend/internal/service/gateway_request.go

@@ -60,6 +60,7 @@ type ParsedRequest struct {
 	Messages        []any           // messages 数组
 	HasSystem       bool            // 是否包含 system 字段（包含 null 也视为显式传入）
 	ThinkingEnabled bool            // 是否开启 thinking（部分平台会影响最终模型名）
+	OutputEffort    string          // output_config.effort（Claude API 的推理强度控制）
 	MaxTokens       int             // max_tokens 值（用于探测请求拦截）
 	SessionContext  *SessionContext // 可选：请求上下文区分因子（nil 时行为不变）
 
@@ -116,6 +117,9 @@ func ParseGatewayRequest(body []byte, protocol string) (*ParsedRequest, error) {
 		parsed.ThinkingEnabled = true
 	}
 
+	// output_config.effort: Claude API 的推理强度控制参数
+	parsed.OutputEffort = strings.TrimSpace(gjson.Get(jsonStr, "output_config.effort").String())
+
 	// max_tokens: 仅接受整数值
 	maxTokensResult := gjson.Get(jsonStr, "max_tokens")
 	if maxTokensResult.Exists() && maxTokensResult.Type == gjson.Number {
@@ -747,6 +751,21 @@ func filterThinkingBlocksInternal(body []byte, _ bool) []byte {
 	return newBody
 }
 
+// NormalizeClaudeOutputEffort normalizes Claude's output_config.effort value.
+// Returns nil for empty or unrecognized values.
+func NormalizeClaudeOutputEffort(raw string) *string {
+	value := strings.ToLower(strings.TrimSpace(raw))
+	if value == "" {
+		return nil
+	}
+	switch value {
+	case "low", "medium", "high", "max":
+		return &value
+	default:
+		return nil
+	}
+}
+
 // =========================
 // Thinking Budget Rectifier
 // =========================

backend/internal/service/gateway_request_test.go

@@ -972,6 +972,76 @@ func BenchmarkParseGatewayRequest_Old_Large(b *testing.B) {
 	}
 }
 
+func TestParseGatewayRequest_OutputEffort(t *testing.T) {
+	tests := []struct {
+		name       string
+		body       string
+		wantEffort string
+	}{
+		{
+			name:       "output_config.effort present",
+			body:       `{"model":"claude-opus-4-6","output_config":{"effort":"medium"},"messages":[]}`,
+			wantEffort: "medium",
+		},
+		{
+			name:       "output_config.effort max",
+			body:       `{"model":"claude-opus-4-6","output_config":{"effort":"max"},"messages":[]}`,
+			wantEffort: "max",
+		},
+		{
+			name:       "output_config without effort",
+			body:       `{"model":"claude-opus-4-6","output_config":{},"messages":[]}`,
+			wantEffort: "",
+		},
+		{
+			name:       "no output_config",
+			body:       `{"model":"claude-opus-4-6","messages":[]}`,
+			wantEffort: "",
+		},
+		{
+			name:       "effort with whitespace trimmed",
+			body:       `{"model":"claude-opus-4-6","output_config":{"effort":" high "},"messages":[]}`,
+			wantEffort: "high",
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			parsed, err := ParseGatewayRequest([]byte(tt.body), "")
+			require.NoError(t, err)
+			require.Equal(t, tt.wantEffort, parsed.OutputEffort)
+		})
+	}
+}
+
+func TestNormalizeClaudeOutputEffort(t *testing.T) {
+	tests := []struct {
+		input string
+		want  *string
+	}{
+		{"low", strPtr("low")},
+		{"medium", strPtr("medium")},
+		{"high", strPtr("high")},
+		{"max", strPtr("max")},
+		{"LOW", strPtr("low")},
+		{"Max", strPtr("max")},
+		{" medium ", strPtr("medium")},
+		{"", nil},
+		{"unknown", nil},
+		{"xhigh", nil},
+	}
+	for _, tt := range tests {
+		t.Run(tt.input, func(t *testing.T) {
+			got := NormalizeClaudeOutputEffort(tt.input)
+			if tt.want == nil {
+				require.Nil(t, got)
+			} else {
+				require.NotNil(t, got)
+				require.Equal(t, *tt.want, *got)
+			}
+		})
+	}
+}
+
 func BenchmarkParseGatewayRequest_New_Large(b *testing.B) {
 	data := buildLargeJSON()
 	b.SetBytes(int64(len(data)))

backend/internal/service/gateway_service.go

@@ -492,6 +492,7 @@ type ForwardResult struct {
 	Duration         time.Duration
 	FirstTokenMs     *int // 首字时间（流式请求）
 	ClientDisconnect bool // 客户端是否在流式传输过程中断开
+	ReasoningEffort  *string
 
 	// 图片生成计费字段（图片生成模型使用）
 	ImageCount int    // 生成的图片数量
@@ -7523,6 +7524,7 @@ func (s *GatewayService) RecordUsage(ctx context.Context, input *RecordUsageInpu
 		AccountID:             account.ID,
 		RequestID:             requestID,
 		Model:                 result.Model,
+		ReasoningEffort:       result.ReasoningEffort,
 		InputTokens:           result.Usage.InputTokens,
 		OutputTokens:          result.Usage.OutputTokens,
 		CacheCreationTokens:   result.Usage.CacheCreationInputTokens,
@@ -7699,6 +7701,7 @@ func (s *GatewayService) RecordUsageWithLongContext(ctx context.Context, input *
 		AccountID:             account.ID,
 		RequestID:             requestID,
 		Model:                 result.Model,
+		ReasoningEffort:       result.ReasoningEffort,
 		InputTokens:           result.Usage.InputTokens,
 		OutputTokens:          result.Usage.OutputTokens,
 		CacheCreationTokens:   result.Usage.CacheCreationInputTokens,

backend/internal/service/openai_gateway_record_usage_test.go

@@ -226,6 +226,41 @@ func TestOpenAIGatewayServiceRecordUsage_UsesUserSpecificGroupRate(t *testing.T)
 	require.Equal(t, 1, userRepo.deductCalls)
 }
 
+func TestOpenAIGatewayServiceRecordUsage_IncludesEndpointMetadata(t *testing.T) {
+	usageRepo := &openAIRecordUsageLogRepoStub{inserted: true}
+	userRepo := &openAIRecordUsageUserRepoStub{}
+	subRepo := &openAIRecordUsageSubRepoStub{}
+	rateRepo := &openAIUserGroupRateRepoStub{}
+	svc := newOpenAIRecordUsageServiceForTest(usageRepo, userRepo, subRepo, rateRepo)
+
+	err := svc.RecordUsage(context.Background(), &OpenAIRecordUsageInput{
+		Result: &OpenAIForwardResult{
+			RequestID: "resp_endpoint_metadata",
+			Usage: OpenAIUsage{
+				InputTokens:  8,
+				OutputTokens: 2,
+			},
+			Model:    "gpt-5.1",
+			Duration: time.Second,
+		},
+		APIKey: &APIKey{
+			ID:    1002,
+			Group: &Group{RateMultiplier: 1},
+		},
+		User:             &User{ID: 2002},
+		Account:          &Account{ID: 3002},
+		InboundEndpoint:  " /v1/chat/completions ",
+		UpstreamEndpoint: " /v1/responses ",
+	})
+
+	require.NoError(t, err)
+	require.NotNil(t, usageRepo.lastLog)
+	require.NotNil(t, usageRepo.lastLog.InboundEndpoint)
+	require.Equal(t, "/v1/chat/completions", *usageRepo.lastLog.InboundEndpoint)
+	require.NotNil(t, usageRepo.lastLog.UpstreamEndpoint)
+	require.Equal(t, "/v1/responses", *usageRepo.lastLog.UpstreamEndpoint)
+}
+
 func TestOpenAIGatewayServiceRecordUsage_FallsBackToGroupDefaultRateOnResolverError(t *testing.T) {
 	groupID := int64(12)
 	groupRate := 1.6

backend/internal/service/openai_gateway_service.go

@@ -4028,6 +4028,8 @@ type OpenAIRecordUsageInput struct {
 	User               *User
 	Account            *Account
 	Subscription       *UserSubscription
+	InboundEndpoint    string
+	UpstreamEndpoint   string
 	UserAgent          string // 请求的 User-Agent
 	IPAddress          string // 请求的客户端 IP 地址
 	RequestPayloadHash string
@@ -4106,6 +4108,8 @@ func (s *OpenAIGatewayService) RecordUsage(ctx context.Context, input *OpenAIRec
 		Model:                 billingModel,
 		ServiceTier:           result.ServiceTier,
 		ReasoningEffort:       result.ReasoningEffort,
+		InboundEndpoint:       optionalTrimmedStringPtr(input.InboundEndpoint),
+		UpstreamEndpoint:      optionalTrimmedStringPtr(input.UpstreamEndpoint),
 		InputTokens:           actualInputTokens,
 		OutputTokens:          result.Usage.OutputTokens,
 		CacheCreationTokens:   result.Usage.CacheCreationInputTokens,
@@ -4125,7 +4129,6 @@ func (s *OpenAIGatewayService) RecordUsage(ctx context.Context, input *OpenAIRec
 		FirstTokenMs:          result.FirstTokenMs,
 		CreatedAt:             time.Now(),
 	}
-
 	// 添加 UserAgent
 	if input.UserAgent != "" {
 		usageLog.UserAgent = &input.UserAgent
@@ -4668,3 +4671,11 @@ func normalizeOpenAIReasoningEffort(raw string) string {
 		return ""
 	}
 }
+
+func optionalTrimmedStringPtr(raw string) *string {
+	trimmed := strings.TrimSpace(raw)
+	if trimmed == "" {
+		return nil
+	}
+	return &trimmed
+}

backend/internal/service/usage_log.go

@@ -100,9 +100,14 @@ type UsageLog struct {
 	Model     string
 	// ServiceTier records the OpenAI service tier used for billing, e.g. "priority" / "flex".
 	ServiceTier *string
-	// ReasoningEffort is the request's reasoning effort level (OpenAI Responses API),
-	// e.g. "low" / "medium" / "high" / "xhigh". Nil means not provided / not applicable.
+	// ReasoningEffort is the request's reasoning effort level.
+	// OpenAI: "low" / "medium" / "high" / "xhigh"; Claude: "low" / "medium" / "high" / "max".
+	// Nil means not provided / not applicable.
 	ReasoningEffort *string
+	// InboundEndpoint is the client-facing API endpoint path, e.g. /v1/chat/completions.
+	InboundEndpoint *string
+	// UpstreamEndpoint is the normalized upstream endpoint path, e.g. /v1/responses.
+	UpstreamEndpoint *string
 
 	GroupID        *int64
 	SubscriptionID *int64