oadmin/sub2api
1
0
Fork 0
Code Issues Pull Requests 1 Actions Packages Projects Releases Wiki Activity
2,804 Commits 2 Branches 0 Tags
1e6912ea2e12beb052d1e0523aa321be619de57f
Commit Graph

1419 Commits

Author SHA1 Message Date
erio
f498eb8fde fix(payment): fix Alipay/Wxpay direct provider type mapping and enable cross-provider load balancing 
Two issues fixed:

1. Alipay.SupportedTypes() returned ["alipay_direct"] and Wxpay returned
   ["wxpay_direct"], but the frontend sends payment_type="alipay"/"wxpay".
   The registry lookup failed with "payment method (alipay) is not
   configured". Fix: return the base types ["alipay"]/["wxpay"].

2. When multiple providers support the same payment type (e.g. EasyPay
   and Alipay direct both handle "alipay"), only the last-registered
   provider's instances were reachable — the registry mapped one type to
   one provider key, and SelectInstance queried by that single key.

   Fix: bypass the registry in invokeProvider and let SelectInstance
   query across all providers when providerKey is empty. The selected
   instance's own ProviderKey (now included in InstanceSelection) is
   used to create the correct provider, enabling true cross-provider
   load balancing.

Closes #1592
 2026-04-13 14:07:12 +08:00
bot
cb016ad861 fix: handle Anthropic credit balance exhausted (400) as account error 
When an Anthropic API key's credit balance is depleted, the upstream
returns HTTP 400 with message containing "credit balance". Previously,
the 400 handler only checked for "organization has been disabled",
so credit-exhausted accounts kept being scheduled — every request
returned the same error.

Treat this case identically to 402 (Payment Required): call
handleAuthError → SetError to stop scheduling the account until
an admin manually recovers it after topping up credits.

Closes #1586
 2026-04-12 13:30:15 +08:00
shuanbao0
422e25c99f fix(gateway): 剥离 Cursor raw body 透传路径中 Codex 不支持的 Responses API 参数 
在前一个 commit 的 isResponsesShape 短路路径基础上,补充对 Cursor 云端
带过来的、Codex 上游统一不支持的顶层 Responses API 参数的剥离:

  - prompt_cache_retention
  - safety_identifier
  - metadata
  - stream_options

根因补充:这条 raw-body 透传路径为了保留 Cursor 的 input 数组整体结构,
不再经过 ChatCompletionsRequest 的反序列化过滤,所以这些 Go 结构体里
没有对应字段的参数会被原样发到上游,上游返回:
    Unsupported parameter: <field>
常规 Chat Completions 转换路径天然通过 ChatCompletionsRequest 丢弃未知字段,
不受影响;此处仅在 isResponsesShape 分支内用 sjson.DeleteBytes 显式过滤,
作用域最小。剥离列表与 openai_gateway_service.go:2034 的
unsupportedFields 语义对齐。

另外在 applyCodexOAuthTransform 的 OAuth 兜底 strip 列表里同步追加
prompt_cache_retention,作为对该函数所有其他 OAuth 调用点的 defense
in depth(当前只有 Cursor 路径的短路已在前面剥过,但保留这一层更稳)。

测试:
- TestCursorMixedShape_StripsUnsupportedFields — 验证所有 4 个字段都被剥
- TestApplyCodexOAuthTransform_StripsPromptCacheRetention — OAuth 兜底路径

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-11 22:48:45 +08:00
shuanbao0
b7edc3ed82 fix(gateway): 兼容 Cursor /v1/chat/completions 的 Responses API body 
Cursor 云端 (User-Agent: Go-http-client/2.0) 发往 /v1/chat/completions 的
body 使用 Responses API 格式:
    {"model":"gpt-5.4","input":[{"role":"system","content":"..."}],"stream":true}

原代码用 ChatCompletionsRequest 反序列化,该结构体没有 Input 字段,
Cursor 的 input 数组被静默丢弃,ChatCompletionsToResponses 转换后产出
input: null,Codex 上游以 "Invalid type for 'input': expected a string,
but got an object" 拒绝请求(上游 typeof null === 'object')。

修复:在 ForwardAsChatCompletions 里用 gjson 检测 body shape,当 input
存在且 messages 缺失时,跳过 Chat→Responses 转换,用 sjson 仅改写 model
字段后原样透传 body。billing 所需的 ServiceTier 和 Reasoning.Effort 通过
gjson 从 raw body 提取,下游 codex OAuth transform 路径保持不变。

测试:新增 openai_cursor_warmup_pipeline_test.go,覆盖 5 个 shape 检测
用例(正向/标准请求不误伤/两字段共存/空 body/JSON 回读)。

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-11 20:22:18 +08:00
erio
fa833f7684 Merge remote-tracking branch 'upstream/main' into feat/payment-system-v2 
# Conflicts:
#	frontend/src/api/admin/settings.ts
#	frontend/src/stores/app.ts
#	frontend/src/types/index.ts
#	frontend/src/views/admin/SettingsView.vue
 2026-04-11 18:25:06 +08:00
erio
e3a000e0d4 refactor(payment): code standards fixes and regression repairs 
Backend:
- Split payment_order.go (546→314 lines) into payment_order_lifecycle.go
- Replace magic strings with constants in factory, easypay, webhook handler
- Add rate limit/validity unit constants in payment_order_lifecycle, payment_service
- Fix critical regression: add PaymentEnabled to GetPublicSettings response
- Add missing migration 099_fix_migrated_purchase_menu_label_icon.sql

Frontend:
- Fix StripePopupView.vue: replace `as any` with typed interface, use extractApiErrorMessage
- Fix AdminOrderTable.vue: replace hardcoded column labels with i18n t() calls
- Fix SubscriptionsView.vue: replace hardcoded Today/Tomorrow with i18n
- Extract duplicate statusBadgeClass/canRefund/formatOrderDateTime to orderUtils.ts
- Add missing i18n keys: common.today, common.tomorrow, payment.orders.orderType/actions
- Remove dead PurchaseSubscriptionView.vue (replaced by PaymentView)
 2026-04-11 13:16:35 +08:00
erio
63d1860dc0 feat(payment): add complete payment system with multi-provider support 
Add a full payment and subscription system supporting EasyPay (Alipay/WeChat),
Stripe, and direct Alipay/WeChat Pay providers with multi-instance load balancing.
 2026-04-11 13:16:35 +08:00
IanShaw027
67a05dfccd fix: honor table defaults and preserve dispatch mappings 2026-04-10 17:55:37 +08:00
IanShaw027
2b70d1d332 merge upstream main into fix/bug-cleanup-main 2026-04-09 21:35:48 +08:00
IanShaw027
b37afd68ec fix(lint): format setting service 2026-04-09 21:31:48 +08:00
Wesley Liddick
bbc79796dc Merge pull request #1529 from IanShaw027/feat/group-messages-dispatch-redo 
feat: 为openai分组增加messages调度模型映射并支持instructions模板注入
 2026-04-09 21:14:38 +08:00
Wesley Liddick
760cc7d6be Merge pull request #1481 from alfadb/fix/increase-error-log-body-limit 
fix(ops): 将错误日志请求体存储限制从 10KB 提升至 256KB
 2026-04-09 21:14:13 +08:00
Wesley Liddick
9a72025afb Merge pull request #1523 from octo-patch/fix/issue-1519-home-content-csp-frame-src 
fix: include home_content URL in CSP frame-src origins
 2026-04-09 21:13:46 +08:00
Wesley Liddick
74302f60ab Merge pull request #1010 from Glorhop/pr/oidc-login 
feat(auth): support OIDC login and prefer IdP real email on sign-in
 2026-04-09 21:13:22 +08:00
IanShaw027
62962c05f1 fix(lint): 修复 CI 中的 ineffassign 和 unused 代码告警，修正 group 排序集成测试兼容性 2026-04-09 19:25:08 +08:00
IanShaw027
5f8e60a1b7 feat(table): 表格排序与搜索改为后端处理 2026-04-09 18:14:28 +08:00
IanShaw027
ad80606a44 feat(settings): 增加全局表格分页配置,支持自定义 2026-04-09 18:14:28 +08:00
alfadb
6401dd7cc7 fix(ops): increase error log request body limit from 10KB to 256KB 
10KB is too aggressive for modern LLM API requests where conversation
context routinely exceeds 1MB. This causes error logs to contain only
a minimal placeholder, making it impossible to debug upstream failures.

256KB retains enough context for effective debugging while the existing
multi-pass trimming logic handles larger payloads gracefully.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-09 17:49:02 +08:00
IanShaw027
7d008bd5b6 fix(test): 修正 admin service 分组测试平台字段赋值 2026-04-09 12:42:37 +08:00
IanShaw027
66ff2def8c fix(test): 补充 admin service 分组测试字符串指针辅助函数 2026-04-09 12:39:05 +08:00
IanShaw027
4de4823a65 feat(openai): 支持messages模型映射与instructions模板注入 2026-04-09 12:29:49 +08:00
IanShaw027
23c4d592f8 feat(group): 增加messages调度模型映射配置 2026-04-09 12:29:28 +08:00
ruiqurm
02a66a01c3 feat: support OIDC login. 2026-04-09 02:20:51 +00:00
octo-patch
ce833d91ce fix: include home_content URL in CSP frame-src origins (fixes #1519) 2026-04-09 09:47:27 +08:00
ius
265687b56d fix: 优化调度快照缓存以避免 Redis 大 MGET 2026-04-08 10:39:15 -07:00
shaw
b982076e52 fix: resolve errcheck lint and add missing enable_cch_signing to test 
- Suppress errcheck for xxhash Digest.Write (never returns error)
- Add enable_cch_signing field to settings API contract test
 2026-04-08 16:23:02 +08:00
shaw
e51c9e50b5 feat: sync billing header cc_version with User-Agent and add opt-in CCH signing 
- Sync cc_version in x-anthropic-billing-header with the fingerprint
  User-Agent version, preserving the message-derived suffix
- Implement xxHash64-based CCH signing to replace the cch=00000
  placeholder with a computed hash
- Add admin toggle (enable_cch_signing) under gateway forwarding settings,
  disabled by default
 2026-04-08 16:11:19 +08:00
Wesley Liddick
5088e91566 Merge pull request #1417 from YanzheL/fix/openai-empty-base64-image-payloads 
fix: sanitize empty base64 image payloads for OpenAI requests
 2026-04-08 14:20:38 +08:00
Wesley Liddick
276f499c82 Merge pull request #1418 from YanzheL/fix/1161-gemini-google-search-grounding 
fix(gemini): preserve google search grounding tools
 2026-04-08 14:19:57 +08:00
Wesley Liddick
5c203ce6c6 Merge pull request #1428 from YanzheL/fix/openai-gateway-content-session-hash-fallback 
fix(gateway): add content-based session hash fallback for non-Codex clients
 2026-04-08 14:17:49 +08:00
Wesley Liddick
47cd1c5286 Merge pull request #1467 from touwaeriol/refactor/channel-service-cleanup 
refactor(channel): split long functions, extract shared validation, move billing validation to service
 2026-04-08 14:16:28 +08:00
Wesley Liddick
06e2756ee4 Merge pull request #1501 from StarryKira/fix/1493-non-streaming-empty-output 
fix: 非流式响应路径扩展SSE检测至所有账号类型
 2026-04-08 14:11:47 +08:00
shaw
1c9a2128cf fix: 修复非CC客户端OAuth伪装被Anthropic检测为第三方应用的问题 
commit f3aa54b 的 rewriteSystemForNonClaudeCode 未能通过 Anthropic 第三方检测，
根因是两个关键信号与真实 Claude Code 不一致：

1. anthropic-beta 头缺少 claude-code-20250219：伪装路径主动将该 beta
   加入 drop set 并移除，但 Anthropic 依赖此 beta 识别 Claude Code 请求。
   修复：非 haiku 模型的伪装请求强制包含 claude-code beta。

2. system 字段使用 string 格式而非 array+cache_control：真实 Claude Code
   始终以 [{type,text,cache_control:{type:"ephemeral"}}] 发送 system，
   string 格式成为第三方检测信号。
   修复：rewriteSystemForNonClaudeCode 改为注入 array 格式。

附带调整：stripSystemCacheControl 按 system 是否被重写动态决定，
重写时保留 CC prompt 的 cache_control，未重写时（haiku/已含CC前缀）
保持原有剥离行为。
 2026-04-08 14:06:06 +08:00
Elysia
9e515ea7c4 fix: 非流式响应路径扩展SSE检测至所有账号类型 (#1493) 
当上游返回SSE格式响应(如sub2api链路)时，API Key账号的非流式路径
未检测SSE，导致终态事件中空output直接透传给客户端。

- 将Content-Type SSE检测从仅OAuth扩展至所有账号类型
- 重命名handleOAuthSSEToJSON为handleSSEToJSON（无OAuth专属逻辑）
- 为透传路径新增handlePassthroughSSEToJSON，支持SSE转JSON及空output重建

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-07 22:49:14 +08:00
Wesley Liddick
00aaf0f796 Merge pull request #1382 from StarryKira/fix/refresh-token-race-condition 
fix: resolve refresh token race condition causing false invalid_grant errors fix issue#1381
 2026-04-07 20:50:08 +08:00
Wesley Liddick
81b96ae123 Merge pull request #1498 from aiexz/main 
do not normalize model for openai API token based accounts
 2026-04-07 20:37:10 +08:00
shaw
7c60ee3c85 feat: Beta策略支持按模型区分处理（模型白名单） 2026-04-07 20:33:09 +08:00
shaw
b2e379cf7a fix: 非流式路径在上游终态事件output为空时从delta事件重建响应内容 
上游API近期更新后，response.completed终态SSE事件的output字段可能为空，
实际内容仅通过response.output_text.delta等增量事件下发。流式路径不受影响，
但chat_completions非流式路径和responses OAuth非流式路径只依赖终态事件的
output，导致返回空响应。

新增BufferedResponseAccumulator累积器，在SSE扫描过程中收集delta事件内容
（文本、function_call、reasoning），当终态output为空时补充重建。

同时修复handleChatBufferedStreamingResponse遗漏response.done事件类型的问题。
 2026-04-07 19:35:56 +08:00
shaw
f3aa54b770 fix: 非Claude Code客户端system prompt迁移至messages以绕过第三方应用检测 
Anthropic近期引入基于system参数内容的第三方应用检测机制，原有的前置追加
Claude Code提示词策略无法通过检测（后续内容仍为非Claude Code格式触发429）。

新策略：对非Claude Code客户端的OAuth/SetupToken账号请求，将system字段
完整替换为Claude Code标识提示词，原始system内容作为user/assistant消息对
注入messages开头，模型仍接收完整指令。

仅影响/v1/messages路径，chat_completions和responses路径保持原有逻辑不变。
真正的Claude Code客户端请求完全不受影响（原样透传）。
 2026-04-07 17:06:47 +08:00
Alex
3a07e92b60 fix(openai): do not normalize /completion API token based accounts 2026-04-07 11:40:41 +03:00
Alex
7eecc49c3a fix(openai): do not normalize API token based accounts 2026-04-07 11:27:57 +03:00
Wesley Liddick
9ab2fd7f9e Merge pull request #1391 from Zqysl/qingyu/fix-openai-passthrough-failover-429-529 
fix(openai): fail over passthrough 429 and 529
 2026-04-07 15:18:52 +08:00
erio
9151d34d40 refactor(channel): split long functions, extract shared validation, move billing validation to service 
- Split Update (98→25 lines), buildCache (54→20 lines), Create (51→25 lines)
  into focused sub-functions: applyUpdateInput, checkGroupConflicts,
  fetchChannelData, populateChannelCache, storeErrorCache, getOldGroupIDs,
  invalidateAuthCacheForGroups
- Extract validateChannelConfig to eliminate duplicated validation calls
  between Create and Update
- Move validatePricingBillingMode from handler to service layer for
  proper separation of concerns
- Add error logging to IsModelRestricted (was silently swallowing errors)
- Add 12 new tests: ToUsageFields, billing mode validation, antigravity
  wildcard mapping isolation, Create/Update mapping conflict integration
 2026-04-05 22:32:49 +08:00
Wesley Liddick
8154575d70 Merge pull request #1464 from touwaeriol/fix/channel-platform-isolation 
fix(channel): remove cross-platform pricing/mapping leakage for antigravity groups
 2026-04-05 21:09:10 +08:00
Wesley Liddick
d757df8a4b Merge pull request #1463 from touwaeriol/feat/remove-sora 
revert: completely remove Sora platform
 2026-04-05 21:08:48 +08:00
erio
c5688fef9a fix: remove cross-platform pricing/mapping leakage for antigravity groups 
Antigravity groups were incorrectly matching pricing and model mapping
entries from anthropic/gemini platform tabs. Each platform should be
strictly isolated — antigravity groups only use antigravity-tagged pricing.
 2026-04-05 20:42:24 +08:00
erio
19655a15f1 fix: gofmt formatting 2026-04-05 18:56:40 +08:00
erio
f345b0f595 fix: use upstream versions of shared files and remove only Sora code 
Restore gateway_service.go, setting_handler.go, routes/admin.go,
dto/settings.go, group_repo.go, api_key_repo.go, wire_gen.go to
upstream/main versions and surgically remove only Sora references.

This preserves upstream-only features (RequireOauthOnly, RequirePrivacySet,
GroupResolution, etc.) that were missing when using release branch versions.
 2026-04-05 18:48:41 +08:00
shaw
f585a15eff fix(billing): prevent channel_mapped override from reverting BillingModel when channel did not map 
When a channel has no model mapping for the requested model, ChannelMappedModel
equals OriginalModel (the user's arbitrary input). Combined with the default
BillingModelSource="channel_mapped", this incorrectly overrides the BillingModel
set by the OpenAI format conversion layer (e.g., gpt-5.4 from DefaultMappedModel)
back to the unmapped original model (e.g., glm) which has no pricing — resulting
in zero-cost billing.

Add guard condition so the channel_mapped override only fires when the channel
actually changed the model (ChannelMappedModel != OriginalModel).
 2026-04-05 17:31:18 +08:00
erio
a29f5a4849 fix: gofmt formatting 2026-04-05 17:22:22 +08:00