i/Aether
1
0
Fork 0
mirror of https://github.com/fawney19/Aether.git synced 2026-10-08 10:27:46 +08:00
Code Issues Packages Projects Releases Wiki Activity
3,058 Commits 6 Branches 182 Tags
ea6b739fd6d39de2a845a7bdafc6bf96abde034f
Commit Graph
1174 Commits
Author SHA1 Message Date
AAEE86 dc3743aecf refactor(ws): 拆分 LogicalTurn 与 ProviderAttempt，结算改表驱动 
评审第 5 条：一个 ResponsesWebSocketTurn 同时代表 logical turn 和 provider
attempt，finalize() 又用 outcome.cancelled() 一个布尔驱动 billing、candidate
状态和供应商效果，于是 provider 终态已经到达、只是最后一跳写客户端失败时，
供应商事实会被 Cancelled 覆盖掉。

- ResponsesWebSocketTurn → ResponsesProviderAttempt，
  ActiveResponsesWebSocketTurn → ActiveProviderAttempt：类型名字明确它只代表
  一次上游执行，logical turn 由 C1 落地的 LogicalTurn 承担。
- 新增 settlement.rs：AttemptProviderOutcome × AttemptClientDelivery 两个正交
  事实，classify_attempt_settlement 一张表推出 status_code / billing /
  candidate 状态 / candidate 错误分类 / 供应商效果 / 是否提交 execution report。
- attempt 观察到 provider 终态即记录 provider_outcome。结算信号
  ResponsesWebSocketTurnOutcome 只回答「为什么现在结算」：ProviderTerminal 与
  Failure 对 provider 是权威的，Cancelled 只描述客户端/连接层面的停止，不再
  覆盖已观察到的 provider 事实。
- candidate 状态与 candidate 错误分类分开输出：现状存在
  「missing_terminal=true 而记账层判 Success」的组合（report kind 不要求观察到
  终态事件时），会写出 status=Success + error_type=stream_missing_terminal_event，
  这个组合必须原样保留。

classify_responses_websocket_turn_effect 的判定表原样搬入 settlement.rs，分支
和顺序均未改动，两个既有不变量测试随之迁移。

行为等价。结算表当前口径与拆分前完全一致：客户端投递失败仍与「供应商声明取消」
落在同一侧（作废账单、candidate 记 Cancelled、只释放 lease、不提交 execution
report），即使 provider 终态已经到达——这一行由
settlement_table_row_client_delivery_failure_currently_voids_a_reached_terminal
锁住现状，修正它是下一步独立的行为修正。

新增 15 个测试：outcome → 双事实映射表逐行（含 stream_timeout 只在 504 失败一族
成立、provider 终态即使 504 也不投射流式超时）、结算表逐行、投递失败时
forced_error 必须为 None、已观察终态不被 Cancelled 覆盖、以及跨整张表的
「每个分支都释放 pool key lease」「作废账单一律不提交 report」不变量。
 2026-08-17 14:53:05 +08:00
AAEE86 1c5ee5228c refactor(ws): 用 ResponsesTurnState 收敛连接 turn 状态 
评审第 2 条：BoundResponsesConnection 用 response_in_flight、active_turn、
active_response_create 三个可独立变化的字段编码同一件事，8 种组合里只有 3 种
合法，非法组合只能靠调用点的 if 和「记得同时改另外两个字段」来避免。

三字段合并为一个 ResponsesTurnState：

  Idle                                  没有进行中的 logical turn
  Responding { logical, attempt }        logical 与 attempt 必须同时存在
  Replanning { logical }                 attempt 已取走去结算/重绑，logical 仍在

Replanning 不是新概念：配额透明重试期间现状就处于这个状态，只是靠
Option::take 意外得到。转换只能走 begin / detach_attempt / resume / end，
response_in_flight 与「是否接受新 response.create」都由变体推导。

由此消除的运行时不变量（原来全靠调用点自觉）：
- 有 attempt 必有 logical turn
- response_in_flight 与 attempt 同生共死（原来 client 写失败后
  active_turn=None 而 response_in_flight 仍为 true）
- logical turn 结束时必须清 attempt：原来 `active_response_create = None`
  在 connection.rs 里手写 13 处，漏一处就残留；现在只有 end() 一个出口
- 上游绑定返回的连接不再自带 response_in_flight=true 的半成品状态

同时删除 update_response_in_flight：Started 帧把已经是 true 的字段再设一次，
Close 帧因为没有解析出的 frame 而根本不触发，是纯冗余写；它在 Idle 态收到
Started 帧时还会把 response_in_flight 置真，从而永久阻塞后续 response.create。

行为等价。ActiveResponsesWebSocketRequest 改名 LogicalTurn 并随状态机移入
新的 turn_state.rs；状态机对 attempt 类型泛型化，测试用轻量替身驱动同一套
转换逻辑，无需 AppState 或真实 socket。
 2026-08-17 14:52:57 +08:00
AAEE86 9d80281b53 fix(ws): route WS planning and continuation through PII redaction 2026-08-17 14:52:49 +08:00
AAEE86 3b036299d4 fix(ws): enforce absolute upstream handshake and initial-message deadlines 2026-08-17 14:52:39 +08:00
AAEE86 f70ae68273 fix(ws): treat max_output_tokens incomplete as legitimate terminal 2026-08-17 14:52:33 +08:00
AAEE86 a498875591 feat(gateway): Responses WebSocket 连通性探针 
新增 aether-codex-ws-probe 与 aether-openai-responses-ws-probe 两个
二进制，用于在不暴露凭据的前提下验证上游 WebSocket 端点可用性：凭据
只从环境变量读取，不写入日志。公共流程放在
bin/support/responses_ws_probe.rs，各 profile 只负责自己的鉴权与
请求头要求。
 2026-08-17 14:51:18 +08:00
AAEE86 71b54070e8 feat(gateway): Codex/OpenAI Responses WebSocket 代理模式 
在 /v1/responses 上支持 WebSocket 升级，把客户端帧中继到上游 Codex /
OpenAI Responses WebSocket 端点，同时保持既有的路由、鉴权、配额与用量
语义：

- 路由与准入：control/route/ai.rs 识别 WebSocket 升级请求；
  websocket/ingress.rs 复用 API Key 鉴权、IP 规则与并发许可，并引入
  独立的 WebSocket 连接许可
- 中继：websocket/responses/* 按 connection / session / turn 分层，
  帧解析归一化、socket 写入有界、continuation 保持调度亲和性
- 配额：orchestration/codex_quota_breaker.rs 在账号配额耗尽时熔断并
  自动恢复，不再直接断开客户端连接
- 用量：每个 turn 的终态用量落库，request_metadata 记录
  websocket_mode / websocket_transport，管理端与 usage 视图暴露
  is_websocket
- 管理端：provider 可配置 Responses WebSocket 开关
 2026-08-17 14:50:33 +08:00
ZheFox 32944538e9 fix(gateway): stop candidate persistence retry storms 2026-08-17 13:49:12 +08:00
ZheFox b13d9b9b40 fix(codex): restore upstream model discovery 2026-08-15 19:36:28 +08:00
zhefox 810c3dfe2b fix(codex): serve versioned dynamic model catalogs 2026-08-14 18:41:44 +08:00
elky 5b0c763086 fix(codex): fence concurrent quota updates 2026-08-14 09:28:07 +08:00
elky f3a12c1008 fix(ai): preserve Codex image edit validation 2026-08-13 11:31:17 +08:00
elky 8cf381b0c3 feat(codex): add OAuth fingerprint convergence 2026-08-13 09:57:17 +08:00
elky edb8362adc fix(provider): omit default model test temperature 2026-08-12 16:56:18 +08:00
elky 29fa4aed19 perf(gateway): raise default server pool floor 2026-08-12 16:56:18 +08:00
zhefox 985ff3c36a test(gateway): align claude_code endpoint reconciliation 2026-08-05 14:34:20 +08:00
zhefox 4d67569873 fix(gateway): support claude_code cross-format Claude messages 2026-08-05 13:57:20 +08:00
zhefox 669f4bddc5 fix: align Responses routing and model permissions 2026-08-03 18:48:01 +08:00
elky 06f5d3c8c0 fix(gateway): complete worker registration cleanup 2026-07-31 11:32:07 +08:00
elky 082407fa51 Merge PR #697: prevent duplicate worker registrations 2026-07-31 11:11:14 +08:00
fawney19 6688ee26db Merge pull request #702 from MMEXA/fix/reconcile-auth-channel-mismatch-formats 
fix(gateway): 修复批量更新 API 格式时的认证通道状态冲突
 2026-07-31 10:28:37 +08:00
elky beb003b7ad feat(models): add external catalog proxy selection 2026-07-31 09:32:25 +08:00
MMEXA 6ecfe0f0a1 fix(gateway): reconcile auth mismatch formats on key update 2026-07-30 22:14:08 +08:00
ZheFox ff47d8d48a fix(gateway): route response history through ai seam 2026-07-30 20:34:41 +08:00
ZheFox 84022c4d48 Merge upstream/main into main 2026-07-30 19:40:39 +08:00
ZheFox 118f441029 feat(gateway): persist OpenAI Responses continuation history 2026-07-30 19:26:52 +08:00
elky 20399b004d Merge PR #700: fix admin pool batch update body buffering 
Preserve main's failover and usage metadata fixes, restore default tunnel regression coverage, and satisfy current Clippy.
 2026-07-30 17:56:37 +08:00
elky 050eb77508 fix(ai): harden responses replay and failure diagnostics 2026-07-30 17:19:54 +08:00
elky 1ab4f079c9 fix(gateway): restore failover and usage diagnostics 2026-07-30 09:12:11 +08:00
MMEXA d7d8db45ba test(gateway): align tunnel error fixture with failover policy 2026-07-30 06:44:59 +08:00
MMEXA e55793c765 fix(ci): satisfy gateway clippy on upstream baseline 2026-07-30 05:14:34 +08:00
MMEXA d8902ea612 fix(gateway): buffer admin pool batch update bodies 2026-07-30 05:14:34 +08:00
elky a04673a90d feat(gateway): harden failover and payload handling 
Retry pre-response transport failures across candidates with an explicit stop policy, and propagate end-to-end timing into usage records and UI diagnostics.

Remove legacy body, import, cookie, PII, and tunnel replay caps while preserving optional operator-configured gateway limits.
 2026-07-30 01:03:27 +08:00
zhefox f8000012f7 fix(ci): stabilize cross-platform workflow checks 2026-07-29 21:55:43 +08:00
worker-2 6080f8cc88 fix(gateway): stabilize worker task records 
Key worker boot records by task so process restarts update the existing
row instead of registering another row for each gateway instance.

Closes #693
Confidence: high
Scope-risk: narrow
 2026-07-29 17:27:51 +08:00
ZheFox 9d9316e434 Merge branch 'fawney19:main' into main 2026-07-28 13:56:34 +08:00
ZheFox 1b697b1111 feat(providers): support FedRAMP Codex agent identity registration 2026-07-28 13:29:30 +08:00
ZheFox f0f87b56a3 feat(providers): add credential-fenced OAuth key cleanup 2026-07-28 11:32:11 +08:00
elky 4148ab1931 fix(routing): harden routed pool scheduling 2026-07-27 22:06:28 +08:00
elky 550cc36760 feat(providers): expand OAuth account management 
Add Claude Code manual and cookie authorization, including redacted batch tasks. Harden OAuth imports, duplicate replacement, provider dialogs, and related account-management tests.
 2026-07-27 15:53:28 +08:00
elky 531cf11025 feat(gateway): harden provider request execution 
Preserve exact request payloads and model client surface and API operation explicitly.

Add Anthropic compatibility profiles, bounded stream commitment, and scoped OAuth retry behavior across provider transports.
 2026-07-27 09:36:31 +08:00
elky 10d369f59c feat(providers): add provider transfer limits 2026-07-26 15:06:56 +08:00
elky 778cfb1a5c feat(data): complete portable SQL backend parity 
Align MySQL and SQLite schemas, migrations, usage, stats, export, and backfill behavior with the shared data contracts. Extend gateway startup and maintenance support across all SQL drivers.
 2026-07-25 21:28:21 +08:00
ZheFox ec36cfbf75 Merge pull request #690 from zhefox/main 
fix(provider): classify deleted Codex agent runtime as invalid
 2026-07-23 11:22:30 +08:00
ZheFox 7bf228a33c fix(provider): classify deleted Codex agent runtime as invalid 2026-07-23 11:21:55 +08:00
elky 3606290ac8 fix(provider): harden Agent Identity OAuth lifecycle 2026-07-23 09:33:00 +08:00
elky fdbc2607ec feat(provider): add dedicated Codex Agent Identity flow 2026-07-22 20:19:29 +08:00
elky 856605defa fix(model-directives): harden suffix configuration 2026-07-22 14:19:09 +08:00
elky 713010fa0a fix(gateway): restore auth role refresh and Rust checks 
Refresh the resolved user role without bypassing owner group and key policies. Resolve Rust 1.95 Clippy failures and make the pending persistence bound test scheduler-independent.
 2026-07-22 11:25:24 +08:00
ZheFox cd2fbeeead Merge pull request #689 from AAEE86/feat/agent-identity-support 
feat(codex): enroll agent identity from session token
 2026-07-22 10:32:06 +08:00