5519 Commits
Author SHA1 Message Date
xlplbo 95a04c8dff test(apicompat): additional_tools 走 chat 桥的回归用例 2026-09-10 19:50:26 +08:00
xlplbo cd1ee1d1ad fix(openai): WS 接入会话状态改用执行作用域键 2026-09-10 17:50:24 +08:00
xlplbo 9e7c2d713c fix(openai): WS 会话抢占改用执行作用域键 2026-09-10 17:50:24 +08:00
xlplbo 06b61ba96f feat(openai): 解析 codex 线程标识派生 WS 执行作用域 2026-09-10 17:50:23 +08:00
feeeei 4a4b50d807 fix(quota): 平台限额表只保存已配置限额的记录,仪表盘按用量与限额展示平台
user_platform_quotas 只保存至少配置了一档限额的记录;不存在的行等价于不限额。

- 注册播种只为配置了限额的平台建行,一个都没有时不写入
- IncrementUsageWithReset 无行时不再 fail-open 建行
- BatchSnapshotUsage 改为只更新既有活跃行,不再插入无限额的新行
- 仓储 BulkInsertInitial/UpsertForUser 把三档全空的记录视为未配置;
  管理员全量替换接口同样剔除,清空限额即软删并放弃累计用量
- 迁移 238 删除三档全 NULL 的历史行
- 用户仪表盘平台卡片 = 有用量的平台 ∪ 配置了限额的平台,计数改为卡片数
- 管理员限额弹窗对未保存限额的平台禁用"重置窗口"
2026-09-10 16:41:58 +08:00
r266-tech 1c0932e173 feat(gateway): retrieve models from the visible catalogue 2026-09-10 15:57:17 +08:00
nekocode a4edda36d5 fix(gateway): 不再剥离客户端 system 上的 cache_control
OAuth mimic 路径无条件删除 system[*].cache_control,不打日志也不报错。
把稳定前缀锚在 system 上的客户端因此拿到 cache_creation 与 cache_read
双 0 —— 不是没命中,是连写都没写。

这个动作是 46e5ac9 引进时「system 必然被整个重写」的配套:客户端的
system 内容都搬进 messages 了,残留断点指着空气,删它是对的。后来
system 注入变成可配置,于是长出了原设计里不存在的组合 —— system 没被
重写,断点照删。剥离失去了赖以成立的前提,代码却还在。

三处调用点里 count_tokens 那处还是硬编码 true,与另两处的
!systemRewritten 语义不一致。

直接移除该选项而不是再加一个开关:注入开启时 system 已被整个替换、
留在上面的断点是我们自己拼进去的稳定锚点;注入关闭时断点就是客户端的
缓存意图。两种情形都没有删它的理由,而需要开关的是有取舍的行为。

顺带补上 count_tokens 的 4 块上限兜底:五条出口里只有它没在自己的转发
路径上调过 enforceCacheControlLimit,原先那句 strip 是这条路上唯一压制
断点数量的东西。拿掉之后「客户端 system + 客户端 messages + 刚注入的
tools[-1]」实测顶到 5 块,而上游对超限是 400。

messages 层的同类改写见 #2369,那次只落地了 messages 一半。

此前该行为零测试覆盖(stripSystemCacheControl 在 _test.go 里零命中),
本次补 4 个用例,其中上限兜底那条撤掉修复即红。

Claude-Session: https://claude.ai/code/session_01WnTW52pmELGy3g7KPEUXj2
2026-09-10 15:48:17 +08:00
reed 1ed36679bd fix(apicompat): recover Responses text from done and terminal events
The streaming Responses-to-Anthropic converter emitted assistant text only
from output_text.delta, so a stream carrying its answer on
response.output_text.done or in the terminal response.output produced a
protocol-valid message with no text.

Record emitted text per output_text part in a ledger that survives content
block closure, and recover only the suffix a finished payload adds, as #6629
does for function-call arguments. Where the part a payload finishes cannot be
established, decline to recover rather than repeat delivered text: terminal
recovery runs only when nothing was emitted, and a done payload indexed
differently from every delta is ignored.

Give output_text.done its own handler so the shared block-closing helper and
its tool-path caller stay unchanged, and so a payload arriving after
message_stop cannot open a content block past the end of the message.
2026-09-09 18:55:16 -05:00
smyyyyy 4e5632c3e3 fix: validate Codex User-Agent before identity parsing 2026-09-10 03:18:31 +08:00
feeeei 44bc47a3e3 fix(gemini): 原生路径登记 2xx 响应里的带内错误信号
Gemini API-key 原生转发对上游 2xx 只抽 usage 原样透传:finishReason 内容拦截、
promptFeedback.blockReason、事件里的 Google 错误信封、空响应都被当作成功,
ops 错误日志、用户错误页、渠道监控全部没有记录。

- 新增 gemini_response_signal:逐事件判定三类带内信号,流式按优先级登记最高者;
  没有 data 事件时对非 SSE 的兜底正文整段判定,空体登记 EMPTY_STREAM / EMPTY_RESPONSE
- 内容策略类按请求级结果登记(invalid_request_error / 400,business limited,
  不做上游归因、不计 SLA);错误信封、生成异常、空响应按上游失败登记并计 SLA
- OpsStreamError 新增 RequestScoped / NonStream:请求级带内行不受残留上游上下文与
  skip_monitoring 影响,落库状态取 IntendedStatus;中间件保留此前尝试的恢复行
- 客户端字节、wire 状态、用量与计费不变
2026-09-10 02:59:15 +08:00
aofee d4e8ddb0c5 fix(grok): allow quota queries during account cooldowns 2026-09-09 21:09:29 +08:00
chihao.ou 21323c31f8 fix(monitor): support upstream base URL paths 2026-09-09 20:02:25 +08:00
wangjunwei de76baaa0d fix(scheduler): preserve Anthropic threshold metadata 2026-09-09 17:13:37 +08:00
刘世俊 a0babc93dd fix(openai): accept full heartbeat bootstrap envelopes 2026-09-09 12:48:57 +08:00
hansnow 7ccc8a6f50 fix(openai): 支持 Image 2.5 并修复 OAuth 生图主控模型 2026-09-09 11:04:59 +08:00
lcy 29a36a10be Merge upstream/main into feat/antigravity-gemini-3.8-flash
Resolve comment-wording conflict in normalizeGeminiThinkingTierAlias by
keeping upstream's phrasing. Function body and behavior are unchanged.
2026-09-09 01:49:32 +08:00
reallyfeishi f180fbb884 fix: anchor channel monitor buckets to UTC 2026-09-08 23:02:06 +08:00
chihao.ou 1d4e0436e4 feat(settings): allow hiding the custom page open button 2026-09-08 21:18:56 +08:00
Wesley Liddick d9efbf8229 Merge pull request #6839 from mason0510/fix/grok-external-web-access
fix(grok): 移除 Chat Completions 中不支持的 external_web_access 参数
2026-09-08 20:47:25 +08:00
Wesley Liddick 233f3dffe6 Merge pull request #6281 from wuji-labs/feat/http2-long-stream-keepalive
feat(upstream): add HTTP/2 PING keepalive for long LLM generation streams
2026-09-08 20:47:07 +08:00
Wesley Liddick c54897a59d Merge pull request #6815 from Pluviobyte/codex/fix-directed-proxy-backups
fix(proxies): make backup references directed and shareable
2026-09-08 20:46:32 +08:00
Wesley Liddick 29646ba6b4 Merge pull request #6488 from aofee/codex/feat-grok-media-eligibility-ui
feat: add Grok media eligibility controls
2026-09-08 20:46:20 +08:00
Wesley Liddick 6f0d0ababc Merge pull request #6320 from spongehah/fix/runtime-block-honor-persisted-cooldown
fix(schedule): 选号按持久化冷却条件清除本机账号级阻断
2026-09-08 20:46:07 +08:00
Wesley Liddick 52f7bcaedd Merge pull request #6759 from ssharkkky/feat/cmv2-user-ranking-toggle
feat(channel-monitor-v2): add user ranking toggle and keep missing TTFT neutral
2026-09-08 20:44:07 +08:00
Wesley Liddick 95acbf1f03 Merge pull request #6424 from cyhhao/fix/bound-ops-system-log-storage
fix(ops): bound database-backed system log growth
2026-09-08 20:43:54 +08:00
Wesley Liddick 43569bb44c Merge pull request #6434 from swjturay/codex/upstream-client-disconnect-drain
fix(openai-ws): drain upstream after client disconnect
2026-09-08 20:42:59 +08:00
Wesley Liddick 15557e7796 Merge pull request #6235 from StarryKira/codex/fix-6232-channel-cache-invalidation
fix: broadcast channel cache invalidations across instances
2026-09-08 20:42:45 +08:00
Wesley Liddick e2fd418a96 Merge pull request #6372 from clansty/fix/openai-oauth-429-fallback-disabled
fix: 修复关闭 429 默认回避后仍触发 OAuth 运行时冷却
2026-09-08 20:42:35 +08:00
Wesley Liddick 4e9b01fd59 Merge pull request #6838 from atogumo/fix/claude-code-probe-any-model
fix(gateway): accept Claude Code max_tokens=1 probes for any model
2026-09-08 20:42:24 +08:00
Wesley Liddick aa167a25a7 Merge pull request #6775 from xlplbo/fix/plugin-zip-handle-windows
fix(plugin): 提交插件包前先关闭 ZIP 读取器,修正 Windows 下两例单测
2026-09-08 20:12:42 +08:00
Wesley Liddick 1e6114c270 Merge pull request #6812 from Pluviobyte/codex/fix-model-plaza-subscriptions
fix(model-plaza): include active subscription groups
2026-09-08 20:10:42 +08:00
Wesley Liddick 9e7039ebca Merge pull request #6811 from Pluviobyte/codex/fix-proxy-partial-update
fix(proxies): preserve omitted expiry and fallback settings
2026-09-08 20:10:20 +08:00
Wesley Liddick 1923d1c27d Merge pull request #6814 from feeeei/fix/go-redis-pool-nil-ctx-panic
fix(redis): 升级 go-redis 至 v9.22.0,修复连接池 nil-ctx panic 竞态
2026-09-08 20:09:59 +08:00
Wesley Liddick b638445234 Merge pull request #6754 from ccemoji/codex/minimal-upstream-fix
fix: cancel streaming upstream before closing body
2026-09-08 20:09:37 +08:00
Wesley Liddick 979fe247f9 Merge pull request #6821 from 22zyfeng-collab/fix/antigravity-oauth-plan-type
fix(antigravity): preserve OAuth plan_type when saving accounts
2026-09-08 20:09:12 +08:00
Wesley Liddick 4892b8f17e Merge pull request #6810 from Pluviobyte/codex/fix-announcement-proxy-date-range
fix: validate announcement and proxy date ranges
2026-09-08 20:08:24 +08:00
Wesley Liddick 28b807a735 Merge pull request #6816 from Pluviobyte/codex/fix-repeated-proxy-fallback
fix(proxies): allow repeated expiry fallback and preserve origin
2026-09-08 20:07:08 +08:00
atogumo 4e5d67df35 fix(gateway): accept Claude Code max_tokens=1 probes for any model
The claude_code_only probe exemption only fired for models containing
"haiku". Claude Code also sends max_tokens=1 requests without a system
prompt to the model currently in use, e.g. when switching models or
refreshing the context usage, so on sonnet/opus groups those requests failed
the system-prompt check and were rejected as non-Claude-Code clients.

Recognise the probe from the request body itself (max_tokens == 1) once the
User-Agent check has passed, regardless of model, and carry max_tokens into
the ParsedRequest projection so the hot path sees it too. The haiku context
flag and the local haiku interception are unchanged; non-haiku probes reach
the pool as ordinary one-token requests. One output token offers nothing to
a client that already forges the CLI User-Agent, so this widens no attack
surface.

Fixes #6591
2026-09-08 19:12:37 +08:00
Wesley Liddick dbe92a1c24 Merge pull request #6758 from DuskLin/feat/minimax-platform
feat(minimax): 增加 MiniMax 平台支持
2026-09-08 18:50:06 +08:00
刘佳林 10940ecf39 fix(minimax): gofmt MiniMax allowlist comments
golangci-lint gofmt failed on aligned comments in the default upstream host list.
2026-09-08 17:49:11 +08:00
刘佳林 3495635a52 fix(minimax): 补齐阈值停调与额度监控接线
调度阈值评估和渠道监控 quota 取数补上 MiniMax 分支;GetCodingPlanProvider 改为只认官方域名,避免把中转 Key 发到厂商额度端点。composite /v1/models 与 Codex 目录纳入 MiniMax;迁移改名为 237;allowlist 去掉未使用的 www.* / api.minimax.cn,abab 前缀收窄到 5/6/7。
2026-09-08 17:41:29 +08:00
mason0510 c8deeb0b05 fix(grok): 移除 Chat Completions 中不支持的 external_web_access 参数
模型:Claude Sonnet 5
修复内容:在 Grok Chat Completions raw 转发流程中添加 external_web_access 参数过滤,修复 Grok API 返回 400 "Argument not supported: external_web_access" 错误
修改理由:Grok xAI API 不支持 external_web_access 参数,Sub2API Responses API 路径已有过滤,但 Chat Completions 路径缺失,导致 grok-4.5 等模型调用失败返回 502
验证结果:代码已通过本地 Demo 验证递归删除逻辑正常,待远端编译部署后验证真实 API 调用
影响范围:仅影响 Grok 平台 Chat Completions API 请求体处理,不影响 Responses API 和其他平台
回滚方式:git revert 本次提交,或手动移除 openai_gateway_chat_completions_raw.go line 159-163 的 sanitizeGrokUnsupportedFields 调用

[AI: 梁心语/Claude Sonnet 5 @ 2026-09-08T17:30]
2026-09-08 16:41:09 +08:00
22zyfeng-collab b6ee9f0a32 fix(antigravity): preserve OAuth subscription plan across account creation 2026-09-08 15:02:32 +08:00
Rain 7137fae2a6 test(proxies): preserve incoming references when editing backups 2026-09-08 13:58:34 +08:00
Rain 859de4c8fd fix(proxies): allow repeated expiry fallback and preserve origin 2026-09-08 13:56:20 +08:00
Rain e835f79819 fix(proxies): make backup references directed and shareable 2026-09-08 13:56:19 +08:00
feeeei 99eba19ffc fix: 升级 go-redis 至 v9.22.0 修复连接池 nil-ctx panic
上游竞态:调用方超时后 wantConn.ctx 被置 nil,拨号 goroutine 用 nil ctx
创建超时上下文触发 panic(go-redis#3630,v9.18.0 起修复)。
生产环境在连接池饥饿 + 请求超时风暴下每秒触发上万次。

同步将 ZRangeByScore 迁移至 ZRangeArgs:v9.22.0 将其标记为 deprecated,
golangci-lint staticcheck 拦截(SA1019)。ZRangeArgs{ByScore: true} 与原
调用发出的 Redis 命令参数完全等价(LIMIT 编码逻辑相同)。
2026-09-08 13:46:59 +08:00
Rain 38cfd7e2d9 fix(proxies): preserve omitted expiry and fallback settings 2026-09-08 13:38:41 +08:00
Rain 3fc08745a6 fix(model-plaza): include active subscription groups 2026-09-08 13:37:53 +08:00
Rain efc6e4a816 fix: validate announcement and proxy date ranges 2026-09-08 13:37:52 +08:00