github/sub2api
4
0
Fork 0
mirror of https://github.com/Wei-Shaw/sub2api.git synced 2026-10-29 08:11:32 +00:00
Code Issues Packages Projects Releases Wiki Activity
6,422 Commits 49 Branches 65 Tags
200b1406d7eb42115ea16ca4fc5363bee9d2155f
Commit Graph
6422 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
Qi HU f4e3eb1c5a fix(openai): detect cyber policy in ws v2 passthrough 2026-08-27 21:50:53 +08:00
Whisper-stark fd872550d5 fix(grok): handle typed invalid tool unions 2026-08-27 21:03:23 +08:00
Whisper-stark f4820c00dc fix(grok): simplify invalid tool union roots 2026-08-27 19:24:50 +08:00
Whisper-stark de6ef7134d fix(grok): sanitize Codex Responses requests 2026-08-27 19:24:50 +08:00
Ryan ChouandClaude Opus 5 e6ea7b9af0 fix(openai): 图像能力丢失时冷却 image 调度，不再反复选中坏号 
OAuth 账号被上游收走 image_generation 能力后，/v1/images/* 稳定返回同一句 400
（Tool choice 'image_generation' not found in 'tools' parameter），但网关不做
任何处置，坏号持续被排程选中（报告者实测连续 17 次同 body 400、窗口内 0 成功、
持续 3.5 小时，直到人工停调度）。

RateLimitService 的 400 分支是白名单式，只认 organization has been disabled /
credit balance / identity verification is required 三句，其余 400 明确不处理；
唯一能兜任意 400 的 tryTempUnschedulable 又要求账号开启 temp-unschedulable
开关，OAuth 账号默认没开。两条路都不命中，shouldDisable 恒为 false，也没有任何
model 作用域冷却。

修复收窄到 sub2api 自建 body 的 images 管线：buildOpenAIImagesResponsesRequest
产出的 body 必定同时带 tool_choice 与匹配的 image_generation tool，这条路上出现
该 400 只可能是上游收了账号能力。forwardOpenAIImagesOAuth 在 upstreamCtx 上打
自建标记（覆盖其下游全部三个错误分支），共用错误入口读到标记才把该 400 判为能力
丢失，对 openai:image_generation 作用域冷却 30 分钟（对齐 upstreamModelNotFound
Cooldown），不禁用整颗账号——账号对文本请求依然健康。

未采用 issue 建议①的全局签名匹配：handleOpenAIAccountUpstreamError 同时服务
passthrough 路径，客户端可自带「tool_choice 指 image_generation 但 tools 不带」
的请求触发一字不差的同一句 400，全局匹配等于任何 API 用户一句话即可禁用健康
账号。附 5 个测试，其中防毒杀用例在撤掉该闸门后会实测失败。建议②（连续 N 次
400 熔断）涉及跨请求状态与配置项，另议。

Fixes #4466

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
 2026-08-27 16:36:01 +08:00
WUJI-Labs 9e7aff59df fix(update): strip hyphenated suffix from version string during comparison 
When running a custom build or pre-release version (e.g. v0.1.183-custom),
parseVersion failed to parse the patch segment because of the non-numeric suffix,
causing compareVersions to incorrectly report an available update.

This commit strips any hyphenated suffix before splitting SemVer components.

Signed-off-by: WUJI-Labs <zoujincool@gmail.com>
 2026-08-27 15:54:22 +08:00
kk 00efee4303 🐛 fix(grok): 为 Grok 4.6 广告 xhigh 推理档位 
Grok Build 的 /v1/models 原先给所有可配 reasoning 的 Grok 模型写死三档，
导致 grok-4.6 无法在 /model 里选出官方支持的 xhigh。转发与 Codex catalog
已按模型判断，现与同一条件对齐。
 2026-08-27 15:50:20 +08:00
zhaotao1 00b7c855cc fix(openai): preserve declared API-key namespace calls 2026-08-27 15:22:43 +08:00
li 4a02e54171 Merge remote-tracking branch 'origin/main' into fix/anthropic-chat-bridge-reasoning-passback 2026-08-27 13:40:29 +08:00
li 9192426d2e merge: 合入 origin/main，与新增的 service_tier 字段并存 
主线在 ResponsesResponse 与两处流式事件上新增了 service_tier；本分支新增的
created_at 与之并列，无语义冲突。同时跟进 ChatCompletionsResponseToResponses
新增的 functionTools 形参。
 2026-08-27 13:40:07 +08:00
li d077002eba fix(openai): 图片工具冷却不再由「模型回文字」这一请求级判据触发 
v0.1.181 起，/v1/images/generations 只要上游回了文字而没回图，就给账号写 30 分钟
openai:image_generation 模型级冷却。该判据由 openAIImagesTextFallbackErrorForText
从模型的自然语言输出**推断**而来，只说明这个 prompt 这一轮模型选择了说话——
仓库自己的既有用例 fixture 就是一句「Here's a polished image prompt for your request.」。

更糟的是同一个错误被判为可重试（502 → IsOpenAIImagesRetryableUpstreamError）并驱动
newOpenAIAccountFailoverError：一次闲聊回复会被逐个账号重试，沿途把每个账号都冷却掉，
表现为「开启后没多久账号就全部被限制」。

改为只在上游 error 帧点名该状态时才写冷却；网关自行推断的判据打上
SynthesizedFromModelText 标记并跳过。换号行为一字未动。

依据是仓库既有政策（alpha/search 路径原话）：工具端点的单次失败
「仍允许本次请求换号，但不修改任何账号状态」。

Fixes #6171
 2026-08-27 13:36:50 +08:00
zenor0 5f09442fce fix(openai): refresh usage after quota reset 2026-08-27 10:09:42 +08:00
feeeei 44003d7f6b fix(gateway): Anthropic/Bedrock 传输层错误转 failover + 持久故障临时摘除账号 
上游 Do/DoWithTLS 返回非 HTTP 错误(代理/DNS/TCP/TLS)时,Anthropic 五条转发路径
(forward/apikey 透传/Bedrock/ForwardAsResponses/ForwardAsChatCompletions)原先
直接向客户端写 502 且不换号,单账号网络故障期间该账号仍持续被调度,请求全量失败。

对齐 OpenAI 侧 handleOpenAIUpstreamTransportError 的既有语义:
- 传输层错误统一返回 *UpstreamFailoverError(502),由 handler failover 循环换号,
  耗尽后才向客户端渲染错误;service 不再写响应
- 持久性故障(connection refused/no route/DNS not found/代理认证失败)额外临时
  摘除账号 10 分钟(仅写库,Anthropic 调度以持久化状态为准,无内存快路径)
- context.Canceled 保持原样返回:不换号也不摘账号
- 分类器改为平台无关命名(classifyUpstreamTransportError)供两侧共用
- Ops 错误事件保留各路径原有字段(UpstreamURL/DurationMs/Passthrough)
 2026-08-27 09:56:41 +08:00
feeeei d881bfc0de fix(openai): 避免 pool 两跳重复计算 system 提示词 2026-08-26 19:55:49 +08:00
wucm667 6ff771d3d3 feat(openai): configure image tool cooldown 2026-08-26 19:07:39 +08:00
Wesley Liddick efb46db0a9 Merge pull request #5926 from baryon/contrib/routed-codex-model-catalog 
feat(gateway): generate complete routed Codex model catalogs
 2026-08-26 15:49:38 +08:00
shaw 8ba81615e0 merge: resolve conflict with main (upstream endpoint observability) 
Merge origin/main into contrib/routed-codex-model-catalog.

Single conflict in openai_gateway_forward.go: main added
ClearActualOpenAIUpstreamEndpoint + SetActualOpenAIUpstreamEndpoint
(commit 4795650d2) at the same insertion point where the PR added
filterOpenAIResponsesNoneReasoningEffortForAccount. Both changes
are independent — endpoint observability first, then body filtering.
 2026-08-26 15:29:01 +08:00
wucm667 50ba146298 fix(openai): preserve multimodal client tool outputs 2026-08-26 15:04:17 +08:00
Wesley Liddick b9083fc7ad Merge pull request #6211 from Monika-shipship/codex/ws-v2-tool-id-regression 
test(openai): 补充 OAuth Responses WS v2 历史 ID 清洗回归测试
 2026-08-26 15:03:24 +08:00
Wesley Liddick 1bf76d26d7 Merge pull request #6229 from Wei-Shaw/fix/openai-upstream-endpoint-observability 
fix(openai): 记录真实上游端点避免错误误报
 2026-08-26 15:03:11 +08:00
Long Li 2abce65031 fix(codex): harden routed catalog capability sync 2026-08-26 12:46:31 +09:00
shaw 6ca1e15b0a chore: update sponsors 2026-08-26 09:58:16 +08:00
Long LiandCursor 195b219707 fix(codex): isolate API-key catalog cache and DeepSeek Codex defaults 
Copy cached API-key manifests before group-specific mutation, use DeepSeek
model IDs for Codex fallbacks, omit unsupported config.toml effort, and
drop wildcard mapping keys from generated catalogs.

Co-authored-by: Cursor <cursoragent@cursor.com>
 2026-08-26 10:56:17 +09:00
shaw 66d664ff08 chore: update sponsors 2026-08-26 09:53:47 +08:00
shaw 4795650d21 fix(openai): record actual upstream endpoint on errors 2026-08-26 09:42:28 +08:00
Rick d522aed65e fix: preserve promo codes for OAuth signup 2026-08-26 08:58:22 +08:00
github-actions[bot] 7634e3c23b chore: sync VERSION to 0.1.183 [skip ci] 2026-08-25 13:53:14 +00:00
Monika d8694f03bc test(openai): cover WSv2 cleanup of stale native tool IDs 2026-08-25 21:45:24 +08:00
Wesley Liddick e8cb019fab Merge pull request #6078 from YogaSakti/fix/responses-custom-tool-call-id 
fix(openai): keep restored tool-call item IDs typed
v0.1.183 		2026-08-25 21:32:34 +08:00
Wesley Liddick 7a7bd3729c Merge pull request #6166 from jyking/fix/email-bind-alias-dedup 
fix(auth): 邮箱换绑增加别名与并发守卫
 2026-08-25 21:31:39 +08:00
Wesley Liddick 49ad7021d7 Merge pull request #6164 from jimbirthday/fix/openai-sticky-capacity-spillover 
fix(openai): preserve sticky binding on capacity spillover
 2026-08-25 21:28:38 +08:00
Wesley Liddick 4952b919a4 Merge pull request #6204 from wucm667/fix/issue-6203-kimi-concurrency-403 
fix(kimi): keep concurrency-limit 403 recoverable
 2026-08-25 21:28:08 +08:00
Wesley Liddick fa0685a4a3 Merge pull request #6175 from wucm667/feat/issue-6172-antigravity-token-clamp 
fix(antigravity): clamp compatible token limits
 2026-08-25 21:27:57 +08:00
Wesley Liddick 8040428716 Merge pull request #6201 from xuhaihan/fix/codex-session-id-affinity 
fix(openai): honor Codex session-id header
 2026-08-25 21:27:24 +08:00
Wesley Liddick 4ff136cfd9 Merge pull request #6193 from antTing/codex/fix-channel-monitor-v2-nullif 
fix(channel-monitor-v2): fix composite platform aggregation SQL
 2026-08-25 21:26:32 +08:00
Wesley Liddick 832cf4df65 Merge pull request #6208 from Wei-Shaw/fix/openai-oauth-429-quota-scheduling 
fix(openai): pause OAuth accounts on exhausted quota 429
 2026-08-25 20:41:41 +08:00
Long LiandCursor 77b6c5bb0c merge: sync contrib/routed-codex-model-catalog with upstream v0.1.182 
Keep catalog membership on schedulable accounts and intersect advertised
capabilities across all active group members that map an alias. Preserve
upstream plugin, service-tier, and models-list body-limit changes.

Co-authored-by: Cursor <cursoragent@cursor.com>
 2026-08-25 21:28:57 +09:00
shaw f1aadd48d5 fix(openai): pause OAuth accounts on exhausted quota 429 2026-08-25 20:25:22 +08:00
Long LiandCursor 5934981e22 test(codex): cover unschedulable catalog capability intersection 
Co-authored-by: Cursor <cursoragent@cursor.com>
 2026-08-25 21:02:47 +09:00
Long LiandCursor db01fb98f8 fix(codex): keep catalog capabilities stable when accounts are unschedulable 
Intersect advertised alias capabilities across all active group members
that map the model, not only currently schedulable accounts, so rate-limits
cannot widen image input or context windows.

Co-authored-by: Cursor <cursoragent@cursor.com>
 2026-08-25 21:01:58 +09:00
wucm667 3802268e2d fix(kimi): keep concurrency 403 recoverable 2026-08-25 19:40:39 +08:00
hank 8e60d57475 fix(openai): honor Codex session-id header 2026-08-25 17:08:27 +08:00
hank 32064d39e7 fix(openai): normalize cross-provider reasoning replay 2026-08-25 17:04:38 +08:00
hank 92a550973a fix(accounts): reauthorize OpenAI with refresh token 2026-08-25 16:52:29 +08:00
暴走的小笼包 b20f29d115 fix(channel-monitor-v2): fix composite platform aggregation SQL 2026-08-25 16:36:04 +08:00
刘佳林 a8cfe746b3 test(usage): cover user and admin reasoning effort page display 
Mount the real usage tables and assert users only see the requested
effort, while admins can reveal mapped effort via column settings.
 2026-08-25 15:49:04 +08:00
刘佳林 5705f4a4aa fix(usage): hide mapped reasoning effort from users 
Users only see the reasoning effort they requested. Admins still see
the requested value plus the forwarded mapped value, matching the
model column's requested vs upstream split.
 2026-08-25 15:36:22 +08:00
刘佳林 11ada80d58 feat(usage): 使用记录展示映射前的推理强度 
用量行新增 requested_reasoning_effort，记录分组策略与模型族改写前的客户端请求值；推理强度列与模型列一致，主行显示请求值，有映射时用 ↳ 展示实际转发值。
 2026-08-25 15:10:52 +08:00
wucm667 1a9898a6a4 fix(antigravity): clamp compatible token limits 2026-08-25 13:13:08 +08:00
github-actions[bot] aa2c4e8d13 chore: sync VERSION to 0.1.182 [skip ci] 2026-08-25 04:47:52 +00:00