Commit Graph

892 Commits

Author SHA1 Message Date
erio
0e69895603 Merge branch 'main' into release/custom-0.1.87
# Conflicts:
#	frontend/src/components/keys/UseKeyModal.vue
2026-02-27 21:20:22 +08:00
erio
3382d496e3 refactor: remove unused detectClaudeMaxCacheBillingOutcomeForUsage function 2026-02-27 20:56:33 +08:00
erio
81d896bf78 fix: sync Antigravity ForwardResult.Usage with client response simulation
Apply Claude Max cache billing to usage before returning ForwardResult
in Antigravity Forward, ensuring RecordUsage gets the same simulated
usage that clients see. Restore apply+fallback in RecordUsage for
consistency across GatewayService and Antigravity paths.
2026-02-27 20:42:53 +08:00
erio
741eae59bb refactor: decouple claude max cache simulation from RecordUsage
Extract setupClaudeMaxStreamingHook and applyClaudeMaxNonStreamingRewrite
facade functions to helpers file. RecordUsage now uses detect-only (no
mutation), client response rewriting handled at Forward layer.
2026-02-27 19:59:36 +08:00
erio
d1b684b782 fix: add 2K image default pricing at 1.5x base price
Previously 2K images used the same base price as 1K ($0.134).
Now 2K uses 1.5x multiplier ($0.201), consistent with 4K using 2x ($0.268).

- Backend: add 2K size branch in getDefaultImagePrice
- Frontend: update 2K placeholder from 0.134 to 0.201
- Tests: update assertions for new 2K default price
2026-02-27 17:37:30 +08:00
erio
b789333b68 test: stabilize subscription progress day assertion 2026-02-27 16:35:37 +08:00
erio
61ef73cb12 refactor: isolate claude max response usage simulation by group toggle 2026-02-27 16:14:07 +08:00
erio
e71be7e0f1 fix: update image pricing tests for 2K tier and refactor claude max billing policy
- Update 5 test assertions to match new 2K default price ($0.201 = base * 1.5)
- Refactor claude max cache billing policy into reusable functions
2026-02-27 15:13:05 +08:00
erio
0302c03864 fix: add 2K image pricing at 1.5x base price 2026-02-27 15:00:18 +08:00
erio
6da2f54e50 refactor: decouple claude max cache policy and add tokenizer 2026-02-27 12:18:22 +08:00
erio
886464b2e9 Merge branch 'feature/claude-max-simulation-review' into release/custom-0.1.86
# Conflicts:
#	backend/cmd/server/VERSION
2026-02-27 09:58:01 +08:00
erio
a6f9f9f968 feat: replace gemini-3-pro-image with gemini-3.1-flash-image
- Add migration 060 to update model_mapping for all antigravity accounts
- Remove gemini-3-pro-image and gemini-3-pro-image-preview mappings
- Add gemini-3.1-flash-image and gemini-3.1-flash-image-preview mappings
- Update frontend usage window to show GImage for new model
- Update isImageGenerationModel to support new model
2026-02-27 09:52:50 +08:00
erio
756b09b6b8 feat: replace gemini-3-pro-image with gemini-3.1-flash-image
- Add migration 060 to update model_mapping for all antigravity accounts
- Remove gemini-3-pro-image and gemini-3-pro-image-preview mappings
- Add gemini-3.1-flash-image and gemini-3.1-flash-image-preview mappings
- Update frontend usage window to show GImage for new model
- Update isImageGenerationModel to support new model
2026-02-27 09:30:44 +08:00
erio
1fb6e9e830 feat: add claude max usage simulation with group switch 2026-02-27 01:54:54 +08:00
alfadb
e6969acb50 fix: address review - fix log wording and add response body assertion in test 2026-02-26 23:49:30 +08:00
alfadb
9489531431 fix(gateway): return 404 instead of fake 200 for unsupported count_tokens endpoint
PR #635 returned HTTP 200 with {"input_tokens": 0} when upstream doesn't
support count_tokens (404). This caused Claude Code CLI to trust the zero
value, believing context uses 0 tokens, so auto-compression never triggers.

Fix: return 404 with proper error body so CLI falls back to its local
tokenizer for accurate estimation. Return nil (not error) to avoid
polluting ops error metrics with expected 404s.

Affected paths:
- Passthrough APIKey accounts: upstream 404 now passed through as 404
- Antigravity accounts: same fix (was also returning fake 200)
2026-02-26 23:34:53 +08:00
shaw
4ac57b4edf fix: 临时移除fast-mode-2026-02-01避免429问题 2026-02-26 15:44:28 +08:00
shaw
c75c6b6858 fix: 将 DriveClient 注入 GeminiOAuthService,消除单元测试中的真实 HTTP 调用
FetchGoogleOneTier 原先在方法内部直接创建 DriveClient 实例,
导致单元测试中对 googleapis.com 发起真实 HTTP 请求,在 CI 环境
产生 401 错误。

将 DriveClient 作为依赖注入到 GeminiOAuthService,遵循项目
端口与适配器架构规范:
- 新增 repository/gemini_drive_client.go 作为 Provider
- 注册到 repository Wire ProviderSet
- 测试中使用 mockDriveClient 替代真实调用
2026-02-26 10:53:04 +08:00
alfadb
03bcd94ae5 fix: count_tokens 端点不支持时降级返回空值 (404 only)
第三方 Anthropic 中转站通常不支持 /v1/messages/count_tokens 端点,
上游返回 404 时降级返回 {input_tokens: 0},客户端 fallback 到本地估算。

- 仅匹配 404 状态码,语义明确:端点不存在
- 其他错误 (400/429/500) 保留原始处理链和 ops 遥测
- 无需解析错误消息内容,不依赖字符串匹配
- 新增 table-driven 测试覆盖 fallback 和 non-fallback 路径
2026-02-26 09:28:45 +08:00
erio
496173da1f merge: 合并上游 v0.1.86 到 main 分支 2026-02-25 19:02:10 +08:00
huangenjun
935ea66681 fix: 修复 sora_sdk_client 类型断言未检查的 errcheck lint 错误
使用安全的 comma-ok 模式替代裸类型断言,避免 golangci-lint errcheck 报错。

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-25 11:43:08 +08:00
huangenjun
65d4ca2563 fix: 修复流式响应中 URL 的 & 被转义为 \u0026 的问题
新增 jsonMarshalRaw 使用 SetEscapeHTML(false) 替代 json.Marshal,
避免 HTML 字符转义导致客户端无法直接使用返回的 URL。

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-25 11:32:56 +08:00
huangenjun
3c619a8da5 refactor: 使用 go-sora2api SDK 替代自建 Sora 客户端
使用 go-sora2api v1.1.0 SDK 替代原有 ~2000 行自建 HTTP/PoW/TLS 指纹代码,
SDK 提供高并发性能优化(实例级 rand、PoW 缓冲区复用、context.Context 支持)。

- 新增 SoraSDKClient 适配器实现 SoraClient 接口
- 精简 sora_client.go 为仅保留接口和类型定义
- 更新 Wire 绑定使用 SoraSDKClient
- 删除 SoraDirectClient、sora_curl_cffi_sidecar、sora_request_guard 等旧代码

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-25 10:15:38 +08:00
Wesley Liddick
1b4e504fad Merge pull request #625 from cagedbird043/pr/antigravity-default-gemini31-passthrough
fix: 默认补全 Antigravity 的 Gemini 3.1 Pro 透传映射
2026-02-25 08:45:16 +08:00
erio
8365a8328b merge: resolve conflicts with upstream/main (Gemini 3→3.1 mappings) 2026-02-25 00:38:39 +08:00
erio
58f21e4b3a fix: correct gofmt alignment in gemini-3.1-pro fallback pricing 2026-02-25 00:23:37 +08:00
erio
5bd7408b2f fix: add fallback pricing for opus-4.6 and gemini-3.1-pro models 2026-02-25 00:10:07 +08:00
erio
c671e8dd1d fix: 统一gemini-3默认映射为非强制3.1 2026-02-24 23:24:48 +08:00
cagedbird043
a3aed3c4c3 fix: 默认补全 antigravity 的 Gemini 3.1 Pro 透传映射 2026-02-24 22:54:11 +08:00
Wesley Liddick
516f8f287c Merge pull request #623 from cagedbird043/fix/antigravity-mapping-upgrade-additions
fix: 补全 Antigravity 模型映射升级与快捷按钮
2026-02-24 22:50:24 +08:00
cagedbird043
ea8104c6a2 fix: antigravity 默认补全 gemini-3-flash 透传 2026-02-24 20:31:36 +08:00
cagedbird043
c10267ce2b fix: 刷新用量成功后自动清理账号可恢复错误状态 2026-02-24 20:04:36 +08:00
erio
da6fd45000 chore: add sonnet-4-6 mapping, config defaults, and CI improvements
- Add claude-sonnet-4-6 to default Antigravity model mapping
- Add antigravity_extra_retries default value in config
- Add cache-dependency-path to CI setup-go for faster builds
- Simplify vitest config to avoid vite plugin compatibility issues
2026-02-24 18:55:39 +08:00
erio
644058174e fix(gemini): enable model_mapping filtering for Gemini API Key accounts
Remove the special case that bypassed model-supported checks for Gemini
API Key accounts, allowing model_mapping to filter requests properly.
Add tests for multiplatform model filtering behavior.
2026-02-24 18:54:59 +08:00
erio
4573868c08 fix(antigravity): bill with mapped model and use final model key for rate limiting
- Use mapped model (billingModel) instead of original request model for billing
- Use resolveFinalAntigravityModelKey for 429 rate limit model key,
  ensuring rate limit records match the actual upstream model
- Add regression tests for both fixes
2026-02-24 18:08:19 +08:00
erio
aaac1aaca9 feat: add mixed-channel precheck API for account-group binding
Add a dedicated CheckMixedChannel endpoint that allows the frontend
to pre-validate mixed channel risk before submitting create/update
requests. This improves UX by showing warnings earlier in the flow
instead of only after form submission.

Backend changes:
- Add CheckMixedChannelRequest struct and CheckMixedChannel handler
- Register POST /check-mixed-channel route
- Expose CheckMixedChannelRisk as public method on AdminService
- Simplify Create/Update 409 responses (remove details/require_confirmation)
- Add comprehensive handler tests and stub methods

Frontend changes:
- Add checkMixedChannelRisk API function and TypeScript types
- Refactor CreateAccountModal to precheck before step transition and submission
- Refactor EditAccountModal to precheck before update submission
- Replace pendingPayload pattern with action-based dialog flow
2026-02-24 17:16:53 +08:00
erio
59898c16c6 fix: fix intercept_warmup_requests config not being saved
Extract applyInterceptWarmup utility to unify all credential building
call sites:
- Fix upstream account creation missing intercept_warmup_requests write
- Fix apikey edit mode missing else-branch to clear the setting
- Add backend unit test for IsInterceptWarmupEnabled
- Add frontend unit test for credentialsBuilder
2026-02-24 16:48:16 +08:00
erio
0dacdf480b fix: distinguish client disconnection from upstream retry failure
Before this change, when a client disconnected mid-request, the error
message was "Upstream request failed after retries", which is misleading
and pollutes error logs. Now we check context.Err() to return a more
accurate "Client disconnected" message for both Claude and Gemini
forward paths.
2026-02-24 16:45:08 +08:00
shaw
e435a46db5 fix: 修复 antigravity UserAgent 重构遗留的编译错误和测试不匹配
- oauth.go: GetUserAgent() 缺少闭合大括号导致语法错误
- client_test.go/oauth_test.go: UserAgent 变量已重构为 GetUserAgent(),更新测试引用
- model_rate_limit_test.go: gemini-3-pro-preview 映射目标已更新为 gemini-3.1-pro-high,同步测试
2026-02-24 14:44:57 +08:00
yangjianbo
c2567831d9 fix(service): 使用 os.Root 修复 Sora 存储路径告警
- 将媒体写入和删除切换为 os.Root 沙箱 API
- 移除旧的路径拼接校验辅助函数并收敛删除逻辑
- 调整并新增相关单元测试覆盖删除行为

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-23 16:06:04 +08:00
yangjianbo
e8671fd7c2 fix(service): 修复 Sora 媒体落地路径穿越风险
- 新增安全路径拼接校验,确保目标文件仍在下载目录内
- 清理失败下载文件时复用安全校验,避免不安全删除路径
- 增加扩展名白名单归一化与相关单元测试

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-23 14:42:07 +08:00
yangjianbo
5fa45f3b8c feat(idempotency): 为关键写接口接入幂等并完善并发容错 2026-02-23 12:45:37 +08:00
erio
0ad20c9489 fix: 修复 Antigravity 账号 intercept_warmup_requests 配置无法保存
提取 applyInterceptWarmup 纯函数统一所有调用点:
- 修复 upstream 创建时遗漏写入 intercept_warmup_requests
- 修复 apikey 编辑时缺少 else 清除逻辑
- 添加前后端单元测试
- 修复 vitest.config.ts mergeConfig 兼容性问题
2026-02-22 23:28:11 +08:00
yangjianbo
7be1195281 feat(api-key): 增加 API Key 上次使用时间并补齐测试 2026-02-22 22:07:17 +08:00
yangjianbo
10636d8a1f fix(codex): 修复额度窗口过期展示并补齐高覆盖测试
- 后端新增绝对重置时间字段计算(codex_5h_reset_at/codex_7d_reset_at)

- 前端统一窗口解析逻辑:绝对时间优先,updated_at+seconds 回退,过期自动归零

- 新增后端与前端单元测试,覆盖关键边界与异常场景
2026-02-22 21:04:52 +08:00
yangjianbo
c67f02eaf0 fix(jwt): 修复仅配置小时时会话提前失效问题
- 将 jwt.access_token_expire_minutes 默认值改为 0,未显式配置时回退 expire_hour

- 调整配置校验为允许 0,仅拒绝负数并补充优先级注释

- 新增配置与认证服务单元测试,覆盖分钟优先与小时回退场景

- 更新示例配置文档,明确分钟/小时优先级与默认行为
2026-02-22 17:37:35 +08:00
yangjianbo
0b32f61062 fix(ratelimit): 清除限流时同步清理临时不可调度状态
- ClearRateLimit 增加清理 temp_unschedulable 与缓存\n- 新增 ClearRateLimit 相关单元测试覆盖成功与失败分支
2026-02-22 17:00:29 +08:00
yangjianbo
2ee6c26676 fix(gateway): 修复粘性会话预取分组错配并优化并发等待热路径 2026-02-22 16:43:33 +08:00
yangjianbo
a89477ddf5 perf(gateway): 优化热点路径并补齐高覆盖测试 2026-02-22 13:31:30 +08:00
yangjianbo
33db7a0fb6 feat(gateway): 引入使用量记录有界 worker 池与自动扩缩容
- 新增 UsageRecordWorkerPool,支持有界队列、溢出降级策略与自动扩缩容
- 将 Gateway/OpenAI/Sora/Gemini 使用量记录改为提交到统一任务池执行
- 增加 usage_record 配置默认值与校验规则,并补充配置与任务提交相关测试
- 注入并托管 worker 池生命周期,服务退出时统一 StopAndWait

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-22 12:56:57 +08:00