GitLab

Anthropic API 的 cache_creation 对象区分了 ephemeral_5m 和 ephemeral_1h
两种缓存创建 token，1h 单价远高于 5m（如 claude-3-5-haiku: 5m=$1/MTok,
1h=$6/MTok）。此前系统统一按 5m 单价计费，导致计费偏低。

后端：
- pricing_service: 加载 LiteLLM 的 cache_creation_input_token_cost_above_1hr
- billing_service: GetModelPricing 启用分类计费（安全守卫 1h>5m），
  CalculateCost 按 5m/1h 分别计费，无明细时回退到 5m 单价
- gateway_service: parseSSEUsage/handleNonStreamingResponse 用 gjson
  提取嵌套 cache_creation 对象的 ephemeral_5m/1h_input_tokens
- antigravity_gateway_service: extractSSEUsage/extractClaudeUsage 同步提取
- usage_log: 修复 GORM column tag 确保写入正确的数据库列
- 新增迁移 054: 删除 GORM 自动生成的重复列

前端：
- 使用记录 tooltip 展示 5m/1h 缓存创建明细（带彩色 badge 区分）
- 表格单元格缓存写入数值旁显示 1h 标识

当账号仅触发 5h 窗口限流时，旧逻辑从聚合头
anthropic-ratelimit-unified-reset 读取重置时间，该值为所有窗口的
最大值（即 7d 重置时间），导致账号被标记为不可调度约 6 天。

新增 calculateAnthropic429ResetTime 函数，解析 Anthropic 的
per-window 头（5h-utilization/reset、7d-utilization/reset、
surpassed-threshold），判断实际触发的窗口并使用对应的重置时间：
- 仅 5h 超标 → 使用 5h-reset（约 5 小时）
- 仅 7d 超标 → 使用 7d-reset
- 两者均超标 → 使用 7d-reset（较长冷却）
- per-window 头不存在 → 回退到聚合头（向后兼容）

- Add groupID parameter to ListAccounts and ListWithFilters methods
- Implement account filtering by group ID in repository query
- Add group query parameter parsing in account handler
- Update all ListAccounts/ListWithFilters call sites with groupID parameter
- Add group filter UI component to AccountTableFilters
- Add i18n translations for group filter label in English and Chinese
- Update API contract and test stubs to reflect new signature
- Enable filtering accounts by their assigned groups in admin panel

- Add used_by_email column to redeem code export CSV for better user identification
- Implement rate_limited status filter in account listing with RateLimitResetAt check
- Extend redeem code search to include user email in addition to code matching
- Add API key search capability to user listing filters
- Display user email in redeem code table used_by column for improved visibility
- Update search placeholders in UI to reflect expanded search capabilities (email, username, notes, API key)
- Improve Chinese and English localization strings for search hints

Gemini 2.5 Pro/Flash thinking models return thoughtsTokenCount separately
from candidatesTokenCount in usageMetadata, but this field was not parsed
or included in billing calculations, causing thinking tokens to be
unbilled.

- Add ThoughtsTokenCount field to GeminiUsageMetadata struct
- Include thoughtsTokenCount in OutputTokens across all 3 Gemini usage
  parsing paths (non-streaming, streaming, compat layer)
- Add tests covering thinking token scenarios

Closes #554

- 新增 OnboardUser API 客户端方法，支持账号 onboarding 获取 project_id
- loadProjectIDWithRetry 增加 onboard 回退：LoadCodeAssist 未返回 project_id 时自动触发 onboarding
- GetAccessToken 中 project_id 补齐改用轻量 FillProjectID 替代全量 RefreshAccountToken
- 补齐逻辑增加 5 分钟冷却机制，防止频繁重试
- OnboardUser 轮询等待改为 context 感知，支持提前取消
- 提取 mergeCredentials 辅助方法消除重复代码
- 新增 extractProjectIDFromOnboardResponse 和 resolveDefaultTierID 单元测试

🧠 refactor(antigravity): 完善 thinking 预算分配策略并重构工具构建逻辑

✨ feat(antigravity): 支持 thinking adaptive 类型并适配 Opus 4.6 动态预算
🧪 test(gateway): 增加 thinking 模式解析与签名块过滤的边界用例测试

后端新增 ValidateRefreshToken service 方法和 POST /oauth/refresh-token 端点，
前端新增 API/Composable/UI 集成，OAuthAuthorizationFlow i18n 动态化，
支持在 Antigravity 创建账号时批量粘贴 Refresh Token 自动验证并创建账号。

- gofmt: 修复 error_passthrough_service.go 格式问题
- errcheck: 修复 error_passthrough_runtime_test.go 类型断言未检查
- staticcheck: if-else 改为 switch (gateway_service.go)
- test: 修复两个测试用例错误使用 MODEL_CAPACITY_EXHAUSTED 导致走错路径

- MatchRule 延迟/限制 body ToLower，先用 statusCode 短路，只在需要关键词匹配时转换且限制 8KB
- 预计算规则的小写关键词/平台和 error code set，消除运行时重复 ToLower 和线性扫描
- MODEL_CAPACITY_EXHAUSTED 全局去重，避免并发请求重复重试同一模型
- 503 重试 body 读取限制从 2MB 降至 8KB
- time.After 替换为 time.NewTimer，防止 context 取消时 timer 泄漏

- ops_error_logger: status < 400 分支增加 OpsSkipPassthroughKey 检查
- ops_upstream_context: 新增 checkSkipMonitoringForUpstreamEvent，中间重试/故障转移事件也能触发跳过标记
- gateway_handler/openai_gateway_handler/gemini_v1beta_handler: handleFailoverExhausted 匹配规则后设置 OpsSkipPassthroughKey
- antigravity_gateway_service: writeMappedClaudeError 增加 applyErrorPassthroughRule 调用

在每条错误透传规则上新增 skip_monitoring 选项，开启后匹配该规则的错误
不会被记录到 ops_error_logs，减少监控噪音。默认关闭，不影响现有规则。

MODEL_CAPACITY_EXHAUSTED (503) 表示模型容量不足，所有账号共享同一容量池，
切换账号无意义。改为固定1s间隔重试最多60次，重试耗尽后直接返回上游错误。

- 新增 antigravityModelCapacityRetryMaxAttempts=60 和 antigravityModelCapacityRetryWait=1s
- shouldTriggerAntigravitySmartRetry 新增 isModelCapacityExhausted 返回值
- handleSmartRetry 对 MODEL_CAPACITY_EXHAUSTED 使用独立重试策略
- handleModelRateLimit 对 MODEL_CAPACITY_EXHAUSTED 仅标记 Handled，不设限流
- 重试耗尽后不设置模型限流、不清除粘性会话、不切换账号

For retryable transient errors (Google 400 "invalid project resource name"
and empty stream responses), retry on the same account up to 2 times
(with 500ms delay) before switching to another account.

- Add RetryableOnSameAccount field to UpstreamFailoverError
- Add same-account retry loop in both Gemini and Claude/OpenAI handler paths
- Move temp-unschedule from service layer to handler layer (only after
  all same-account retries exhausted)
- Reduce temp-unschedule cooldown from 30 minutes to 1 minute

- Empty stream responses now return UpstreamFailoverError instead of
  plain 502, triggering automatic account switching (up to 10 retries)
- Add tempUnscheduleEmptyResponse: accounts returning empty responses
  are temp-unscheduled for 30 minutes
- Apply to both Claude and Gemini non-streaming paths
- Align googleConfigErrorCooldown from 60m to 30m for consistency

Google 后端间歇性返回 400 "Invalid project resource name" 错误，
此前该错误直接透传给客户端且不触发账号切换，导致请求失败。

- 在 Antigravity 和 Gemini 两个平台的所有转发路径中，
  精确匹配该错误消息后触发 failover 自动换号重试
- 命中后将账号临时封禁 1 小时，避免反复调度到同一故障账号
- 提取共享函数 isGoogleProjectConfigError / tempUnscheduleGoogleConfigError
  消除跨 Service 的代码重复

覆盖 Service 层和 Handler 层的所有新增逻辑：
- isSingleAccountRetry context 标记检查
- handleSmartRetry 中 503 + SingleAccountRetry 分支
- handleSingleAccountRetryInPlace 原地重试逻辑
- antigravityRetryLoop 预检查跳过限流
- sleepAntigravitySingleAccountBackoff 固定延迟退避
- 端到端集成场景验证
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Add ShouldHandleErrorCode guard at the entry of handleGeminiUpstreamError
and AntigravityGatewayService.handleUpstreamError so that accounts with
custom error codes (e.g. [599]) are not rate-limited when the upstream
returns a non-matching status (e.g. 429).

When custom error codes are enabled and the upstream error code is NOT
in the configured list, return HTTP 500 to the client instead of
transparently forwarding the original status code.

Also adds integration test TestCustomErrorCode599 verifying that 429,
500, 503, 401, 403 all return 500 without triggering SetRateLimited
or SetError.

单账号 antigravity 分组收到 503 (MODEL_CAPACITY_EXHAUSTED) 时，
原逻辑会设置 ~29s 模型限流标记。由于只有一个账号无法切换，
后续所有新请求在预检查时命中限流 → 几毫秒内直接返回 503，
导致约 30 秒的雪崩窗口。

修复：在 Handler 入口处检查分组是否只有单个 antigravity 账号，
如果是则提前设置 SingleAccountRetry context 标记，让 Service 层
首次 503 就走原地重试逻辑（不设限流标记），避免污染后续请求。

当分组内只有一个可用账号且上游返回 503 (MODEL_CAPACITY_EXHAUSTED) 时，
不再设置模型限流+切换账号（因为切换回来还是同一个账号），而是在 Service 层
原地等待+重试，避免双重等待问题。

主要变更：
- Handler 层：检测单账号 503 场景，清除排除列表并设置 SingleAccountRetry 标记
- Service 层：新增 handleSingleAccountRetryInPlace 原地重试逻辑
- Service 层：预检查跳过单账号模式下的限流检查
- 新增 ctxkey.SingleAccountRetry 上下文标记

- 后端新增 PreviewFromCRS 接口，允许用户先预览 CRS 中的账号
- 后端支持在同步时选择特定账号，不选中的账号将被跳过
- 前端重构 SyncFromCrsModal 为三步向导：输入凭据 → 预览账号 → 执行同步
- 改进表单无障碍性：添加 for/id 关联和 required 属性
- 修复 Back 按钮返回时的状态清理
- 新增 buildSelectedSet 和 shouldCreateAccount 的单元测试
- 完整的向后兼容性：旧客户端不发送 selected_account_ids 时行为不变

- Fix gofmt alignment in ops_realtime_models.go
- Remove SetAntigravityQuotaScopeLimit mock from api_contract_test.go
- Add UpdateSortOrders mock to mockGroupRepoForGateway

Merge functional changes from develop branch:
- Remove AntigravityQuotaScope system (claude/gemini_text/gemini_image)
- Replace with per-model rate limiting using resolveAntigravityModelKey
- Remove model load statistics (IncrModelCallCount/GetModelLoadBatch)
- Simplify account selection to unified priority→load→LRU algorithm
- Remove SetAntigravityQuotaScopeLimit from AccountRepository
- Clean up scope-related UI indicators and API fields

Add post-sort shuffle for accounts with identical (priority, loadRate,
lastUsedAt) to break deterministic ordering when concurrent requests
read the same scheduler snapshot. Applies to both Antigravity and
OpenAI scheduling paths, plus the sortAccountsByPriorityAndLastUsed
helper.

Keeps upstream CallCount/ModelLoadInfo scheduling intact; shuffle is
additive and only randomises within equivalent-rank groups.

Without this routing guard, ForwardUpstream is never called because
Forward() always proceeds with the standard OAuth/cookie flow.

- In handleSmartRetry, use the actual upstream retryDelay to set model
  rate limit duration instead of always using the 30s default
- Return info.RetryDelay from shouldTriggerAntigravitySmartRetry when
  shouldRateLimitModel=true, so callers know the actual delay
- Extract getDefaultRateLimitDuration() and resolveResetTime() helpers
  to reduce duplication in handleUpstreamError 429 handling
- Improve debug logging with upstream_retry_delay and response body