github/sub2api
4
0
Fork 0
mirror of https://github.com/Wei-Shaw/sub2api.git synced 2026-10-28 08:08:32 +00:00
Code Issues Packages Projects Releases Wiki Activity
5,834 Commits 49 Branches 64 Tags
7d633f5fc3dcedea83c77ce77bc25dc497b88e0a
Commit Graph
5834 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
IanShaw027 825f9c78d5 fix(channel-monitor-v2): default mode v1 and gentle adaptive backfill 
Default channel_monitor_mode to v1 (opt-in V2) so upgrades keep active
probes; existing explicit v2 rows are left alone via ON CONFLICT DO NOTHING
plus migration checksum compatibility for already-applied 195.

V2 first-enable backfill no longer compresses ticks to 5s or uses 24h
chunks. Each tick does recent overlap plus at most one historical chunk
with depth-based ceilings (2h/4h/6h), adaptive grow/shrink, and failure
backoff. Error request_id dedup is bounded by a 90-minute lookback so
ops_error_logs is not scanned for full history.

Also align hide_throughput parse default with privacy-preserving public
runtime (missing key → true).
 2026-08-08 13:59:11 +08:00
IanShaw027 cec922d335 fix(grok): clear golangci-lint findings on complete-integration branch 
Check Close/CloseNow errors, drop unused helpers and dead constants,
lowercase ST1005 error strings, and stop discarding unwrap status as an
unused assignment so CI golangci-lint passes.
 2026-08-08 12:56:40 +08:00
IanShaw027 a4faa015a7 fix(deps): bump nanoid 3.3.16 -> 3.3.17 for GHSA-2v37-7h3g-55p8 
Match upstream lockfile pin so frontend-security audit gate passes.
 2026-08-08 12:50:53 +08:00
IanShaw027 3c22aeeb3d fix(channel-monitor-v2): default hide_throughput true for privacy parity 
Align InitializeDefaultSettings and parseSettings with the public-settings
path and migration 206 so admin and user views agree when the key is unset.
 2026-08-08 12:49:39 +08:00
IanShaw027 1f58e25ab3 Merge upstream/main into feat/grok-complete-integration 
冲突集中在 chat completions / messages 两条 Responses 转发路径：
upstream 给 OpenAIForwardResult 增加了 UpstreamResponseModel 与
UpstreamResponseModelConflict（配套 beginUpstreamResponseModelObservation
观测器），本分支在同样位置把返回值改成了具名变量以便挂 Grok 原生搜索计数。
两侧不互斥，合并结果同时保留上游的响应模型观测字段与 Grok SearchCount 逻辑。

frontend/pnpm-lock.yaml 取 upstream 版本：package.json 与 upstream 完全一致，
本地差异只是 pnpm install 的重解析噪音。
 2026-08-08 11:12:56 +08:00
IanShaw027 07b46e93e4 fix(grok): 修正 voice 路由推导、视频价归一化与门禁缓存驻留 
- custom-voices endpoint 改由匹配到的路由模板推导（c.FullPath()）。
  原实现按请求 URL 字面后缀判 /audio，voice_id 恰为 "audio" 时
  GET /custom-voices/audio 会被改写成 custom-voices/audio/audio，
  把档案查询变成音频下载。补 voice_id="audio" 的路由推导测试。

- NormalizeVideoModelPrices 不再把无法识别的分辨率静默折算成 480p：
  新增 LookupVideoBillingResolution 报告未知档位，配置解析路径丢弃并告警，
  运行时计费仍走 OrDefault 兜底。model/tier 两层遍历改为排序遍历，
  多个别名收敛到同一 family 时结果不再随 Go map 顺序漂移，冲突单价告警。

- grok free quota 软门禁缓存新增过期淘汰。条目按 account_id 键控且只写不删，
  账号下线后会驻留至进程结束；淘汰只挂在已受 TTL 约束的查询路径上，
  不影响缓存命中热路径。

- 迁移 220 清空非 Grok 分组视频价前先落快照表 groups_video_price_backup_220，
  原 UPDATE 不可回滚。

- 简化 grok_search_count 两处等价冗余的事件类型分支。
 2026-08-08 11:02:22 +08:00
Wesley Liddick cc67b1aca1 Merge pull request #5406 from bestony/fix/openai-oauth-routing-hints 
fix(openai): forward OAuth routing hints
 2026-08-08 10:58:53 +08:00
shaw 8ad0a5ff52 fix(deps): bump nanoid 3.3.16 -> 3.3.17 to clear GHSA-2v37-7h3g-55p8 audit gate 2026-08-08 10:48:42 +08:00
IanShaw027 b7112d596f fix(keys): grok-4.5 上下文窗口按 500k 对齐并修正账号测试 i18n mock 
- UseKeyModal 中 grok-4.5 的 context_window / model_context_window /
  OpenCode 模型目录统一改为 500000，与 xAI 实际上下文一致
- UseKeyModal.spec 的 OpenCode 目录断言同步为 500000
- AccountTestModal.spec 补齐 imagePreviewAlt 的 i18n mock，
  跟上组件把 alt 文案迁到 i18n 之后的行为
 2026-08-08 10:40:40 +08:00
IanShaw027 b717ed9d0d feat(keys): complete Grok Build sample with endpoints/auth/session/features 
Expand Use Key Grok CLI config.toml to include production-oriented sections
from working gateway setups: full [endpoints], preferred_method=api_key,
image_description, auto_compact threshold, and Imagine image/video feature
overrides with guided comments.
 2026-08-08 10:33:45 +08:00
IanShaw027 f3bac4619e feat(keys): expand Grok client samples and tune free soft-gate default 
Ship Use Key templates that match Grok Build / Codex best practice: env
vars + multi-model config.toml with api_backend=responses, env_key over
hardcoded secrets, and clearer shell/path guidance for Claude/Codex/OpenCode.

Also set free_quota_token_limit default to 500k (24h soft-gate), clean up
personal-dev-only comments, and keep billing test fixtures aligned.
 2026-08-08 10:26:16 +08:00
IanShaw027 2e857b20b8 chore(grok): 从 PR 中移除进度文档与 lockfile 噪音 
- 删除 CHANNEL_MONITOR_V2_DESIGN.md（属于其它分支的设计草稿）
- 删除 backend/docs/GROK_INTEGRATION_PROGRESS.md（开发期进度记录，不属于产物）
- 还原 frontend/pnpm-lock.yaml 至 upstream/main（package.json 无变化，
  差异全部来自本地 pnpm install 重解析）
 2026-08-08 10:00:56 +08:00
IanShaw027 b7863650ec fix(usage): prefer explicit video billing mode over image_count 
Reuse shared getDisplayBillingMode so user usage rows with billing_mode=video
are not mislabeled as image when image_count is non-zero.
 2026-08-08 09:45:12 +08:00
IanShaw027 bdeb13021a fix(admin): i18n account-test media uploads and shared file pickers 
Replace native file-input labels with translated choose-file buttons, localize
upload errors and previews, and fix ImageUpload defaults for zh/en.
 2026-08-08 09:45:12 +08:00
IanShaw027 e01ce90d47 fix(grok): harden voice request ids, video pending, and search pricing alerts 
Mint durable grok_audio/grok_realtime usage ids, avoid CLI headers on api.x.ai
voice, retry video pending store and fail-closed when snapshot is missing without
status duration, align pure-video ImageCount tests, and escalate unset search
price_per_1k to error-level logs.
 2026-08-08 09:45:12 +08:00
IanShaw027 12db0f906a fix(grok): drop account-test ZDR path and align media CLI headers 
Remove optional upload_url / fake connectivity-only success from admin video
tests. Stamp Grok CLI headers only on the CLI proxy so OAuth media against
api.x.ai can complete and preview video like the gateway path.
 2026-08-08 09:45:12 +08:00
IanShaw027 8005335742 fix(channel-monitor-v2): align admin error details dedup 2026-08-08 09:00:15 +08:00
IanShaw027 64d1ebe4a5 fix(channel-monitor-v2): close dedup and migration privacy gaps 2026-08-08 08:59:18 +08:00
IanShaw027 9f3ee38d4e fix(channel-monitor-v2): preserve migration checksums and privacy defaults 2026-08-08 08:55:59 +08:00
IanShaw027 77e39fe608 docs(grok): note custom-voices coverage and free 24h soft-gate scope 
Document voice CRUD/audio surface area and clarify free-window local soft-gate
semantics versus upstream metering.
 2026-08-08 08:48:38 +08:00
IanShaw027 85fb776151 test(frontend): mock getLiveCapability and align rollback timeout 
Stub GroupsView live capability API in unit tests and expect the longer
system rollback request timeout; refresh pnpm-lock after dependency resolve.
 2026-08-08 08:48:38 +08:00
IanShaw027 6345b048d1 style(admin): color paid Grok and OpenAI plan badges 
Keep free muted gray; SuperGrok cyan, Heavy purple, and distinct colors for
OpenAI Plus/Team/Pro so paid subscriptions stand out in the account list.
 2026-08-08 08:48:38 +08:00
IanShaw027 e1d6600eb2 fix(admin): simplify Grok usage UI for free 24h and paid windows 
Show free accounts only the rolling 24h soft-gate bar; paid accounts show
7d/30d and prepaid money. Drop the always-visible manual probe chip from the
usage cell in favor of scheduled recovery and usage load.
 2026-08-08 08:48:38 +08:00
IanShaw027 8399a30417 fix(grok): treat free-usage and billing exhaustion as recoverable 
Classify free-usage bodies during account tests without quarantining content
policy, and mark billing/spending-limit refresh failures as transient so
accounts stay probe-eligible.
 2026-08-08 08:48:38 +08:00
IanShaw027 35faaa6d21 feat(grok): register custom-voices CRUD and audio download gateway routes 
Forward list/get/patch/delete and reference-audio paths with safe path segment
encoding, method passthrough, and empty-body GET/DELETE handling.
 2026-08-08 08:48:38 +08:00
IanShaw027 85b65284ec fix(grok): set async video duration_ms from create accept to done discovery 
Store CreatedAt on pending billing at video create and use wall-clock E2E
latency when status/content first observes official done+video.url, so usage
logs no longer record only the single poll hop.
 2026-08-08 08:48:38 +08:00
IanShaw027 0d98176c59 fix(channel-monitor-v2): correct aggregation privacy and backfill 2026-08-08 01:59:44 +08:00
IanShaw027 165b072908 fix(grok): gateway media/voice routing, models, and status UI polish 
Align gateway Grok media/voice paths and model lists, harden upstream failure
and quota handling, clear non-Grok video generation config migration, and polish
temp-unsched/status indicators with model whitelist updates.
 2026-08-08 01:07:27 +08:00
IanShaw027 2526a04226 fix(admin): empty web-search config on missing setting and reset dialog scroll 
Return a disabled empty web-search-emulation config when the setting key is
absent, and reset BaseDialog body scroll on open for long modals.
 2026-08-08 01:07:27 +08:00
IanShaw027 68faeac837 fix(grok): restore base URL resolution and operator settings wiring 
Honor account GetGrokBaseURLOr policy for official vs custom endpoints,
and wire settings resolution used by responses/chat URL builders.
 2026-08-08 01:07:19 +08:00
IanShaw027 6d632eec45 fix(grok): tighten OAuth SSO flow and hide password login 
Require oauth state/redirect consistency, fail closed on missing proxy,
and remove password login from create/reauth UI (admin-only password path stays off by default).
 2026-08-08 01:07:19 +08:00
IanShaw027 d0767eab9d feat(grok): admin account test modes with real media preview 
Add mode-first connectivity probes for text/image/video/search/tts/stt/realtime,
standalone voice and web-search paths, media upload options, and in-browser
image/audio/video preview (including ZDR-safe b64 images and edit validation).
 2026-08-08 01:07:08 +08:00
wucm667 9763765ebf fix(openai): fail over OAuth image stream errors 2026-08-08 01:05:29 +08:00
cyh 74fcdf3d42 fix(openai): preserve exponential capacity backoff 2026-08-08 01:02:02 +08:00
github-actions[bot] 68d8f122e4 chore: sync VERSION to 0.1.172 [skip ci] 2026-08-07 15:39:12 +00:00
Wesley Liddick 155c494964 Merge pull request #5399 from fengshao1227/fix/responses-anthropic-invalid-content-blocks 
fix(apicompat): Responses→Anthropic 转换不再发出上游会拒收的 content block
v0.1.172 		2026-08-07 23:21:59 +08:00
Wesley Liddick 8991574873 Merge pull request #5345 from puppywang/fix/oauth-pending-account-takeover 
fix(security): block OAuth account takeover via pending exchange
 2026-08-07 23:20:22 +08:00
Wesley Liddick 8f7b0a314d Merge pull request #5398 from Wei-Shaw/fix/openai-capacity-shed-stream-recovery 
fix(gateway): 流内降载错误恢复 pre-output failover 并对客户端改写为可重试错误码
 2026-08-07 23:11:20 +08:00
li 64090de664 fix(apicompat): Responses→Anthropic 转换不再发出上游会拒收的 content block 
convertResponsesInputToAnthropic 的 default 分支把未知 item 的 content 逐字
透传成 Anthropic user 消息，Responses 专有的分片类型会原样进入上游请求体。
最典型的是工具执行后回放的 reasoning item：带 content 数组时，reasoning_text
块直接发给 Anthropic，上游回 400 Request body format invalid，而该 item 会一直
留在会话历史里，导致此后每一轮都继续失败。

同时修正两处会产出空内容消息的路径——Anthropic 拒收空内容消息与空白 text 块：
分片全部不可识别时，user 消息退化成 content:""、assistant 消息退化成单个空
text 块。

改动：
- type=reasoning 显式跳过。Anthropic 无法摄入 OpenAI reasoning：encrypted_content
  不透明，thinking 重放需要上游签发的 signature。Codex 常见形态（只带 summary +
  encrypted_content）本来就会被丢弃，这里让带 content 的形态行为一致。
- default 分支改走 convertResponsesUserToAnthropicContent 白名单转换，保留其中
  可识别的文本/图片，丢弃其余分片。
- user / assistant 分支在转换结果为空内容或纯空白文本时跳过该消息。

Fixes #5329
 2026-08-07 23:09:03 +08:00
Wesley Liddick de904c433e Merge pull request #5396 from Brisbanehuang/codex/upstream-response-model-audit 
feat(usage): 增加上游响应模型审计
 2026-08-07 23:00:46 +08:00
shaw 14a27f1960 test(gateway): 校准 error 帧边界 flush 期望至 pre-output failover 新契约 
可重试类 error 帧按设计不再算客户端输出（保留在 attempt 缓冲中，
为随后的 response.failed 保住 pre-output failover），因此不在自身
边界单独 flush，而是与终止帧一起出站（1 次 flush）。原用例按旧行为
断言 2 次 flush 导致 CI 失败。

同时补充不可重试类（invalid_request）error 帧仍在边界 flush 的对照
用例，锁住两类错误帧的行为区分。
 2026-08-07 22:55:42 +08:00
shaw c33c3208e3 fix(gateway): 流内降载错误恢复 pre-output failover 并对客户端改写为可重试错误码 
上游容量降载的真实序列是「event: error → event: response.failed」。此前
{"type":"error"} 帧被当作首个客户端输出立即 flush，clientOutputStarted 被固化，
随后的 response.failed 永远进不了 pre-output failover 分支，降载错误被原样转发；
Codex CLI 对 server_is_overloaded/slow_down 按闭集判致命，直接终止会话并提示
"Selected model is at capacity. Please try a different model."。

修复：
- 可重试类 error 帧不再算客户端输出，恢复既有的同账号重试+切号链路；
  不可重试类（content_policy/invalid_request 等）维持原样转发，保留上游错误细节
- 必须转发给客户端时（流中途已有输出 / WS 桥接），把 server_is_overloaded、
  slow_down 改写为致命集之外的 server_error，触发 Codex 内置退避重试；
  错误消息原样保留，rate_limit_exceeded 等其他错误码一律不动
- 监控、计费与账号状态判定均基于改写前的原始事件
 2026-08-07 22:42:33 +08:00
白宦成 de349187d9 fix(openai): harden priority routing hints 2026-08-07 21:47:54 +08:00
白宦成 815035fcc9 fix(openai): send OAuth routing hints 2026-08-07 21:47:53 +08:00
白宦成 915cc7e7bd fix(openai): stop injecting legacy beta on OAuth responses 2026-08-07 21:47:53 +08:00
Brisbanehuang db0bff82c7 feat(usage): audit upstream response models 
(cherry picked from commit 839036224f795c8ee5dc6718a2a14372a45eea44)
 2026-08-07 09:40:11 -04:00
Wesley Liddick e88fc52ce6 Merge pull request #5383 from fengshao1227/fix/responses-tool-parameters-null-type 
fix(openai): 修正 Responses 工具 Schema 中显式为 null 的 parameters.type
 2026-08-07 21:06:10 +08:00
Wesley Liddick 045b620d0d Merge pull request #5391 from fengshao1227/fix/codex-plan-gated-image-model-no-cooldown 
fix(ratelimit): 图片模型被 Codex 文本端点拒绝时不再写模型冷却
 2026-08-07 21:05:58 +08:00
Wesley Liddick e79429b03f Merge pull request #5392 from fengshao1227/fix/grok-405-failover-and-cooldown 
fix(grok): 405 纳入 failover 与临时踢号，修复 Grok 会话粘性锁死
 2026-08-07 21:05:46 +08:00
li 02fbcbe3ad fix(ratelimit): 守卫按端点来源门控，并与冷却键对齐模型口径 
上一版守卫只看模型类型，不区分请求从哪个端点进来。OAuth 账号的 /v1/images/*
上游同样是 Codex Responses（openai_images_responses.go → handleOpenAIImagesErrorResponse
→ handleOpenAIAccountUpstreamError → HandleUpstreamModelNotFound），所以专用生图
端点也会命中 plan-gated 分支。账号确实不具备生图能力时跳过冷却，会让调度层失去
唯一的刹车：每个请求都完整走一遍号池，对上游形成无上界的 400 放大。

改动：
- 新增 ctxkey.OpenAIImagesEndpoint 与 WithOpenAIImagesEndpoint /
  OpenAIImagesEndpointFromContext，在 handler/openai_images.go 入口置位；
  与 OpenAIImageGenerationIntent 区分——后者在 /v1/responses 带图片模型时也会置位。
- 守卫下移到 modelKey 计算之后，抽成 shouldSkipCodexPlanGatedImageModelCooldown，
  仅在 plan-gated 分支、且非 /v1/images/* 入站时生效。
- 同时判断 requestedModel 与最终 modelKey：冷却键走 account.GetMappedModel，
  账号可以把文本别名映射到 gpt-image-*，只判请求模型会漏掉这种形态。
 2026-08-07 20:49:39 +08:00