feat(openai): OpenAI Fast/Flex Policy 完整实现（HTTP + WebSocket + Admin）
对称参照 Claude BetaPolicy 的 fast-mode 过滤实现，新增针对 OpenAI 上游
service_tier 字段（priority / flex，含客户端 "fast" → "priority" 归一化）的
pass / filter / block 三态策略，覆盖全部 OpenAI 入口 + admin 配置入口。
后端核心
- 新增 SettingKeyOpenAIFastPolicySettings、OpenAIFastPolicyRule、
OpenAIFastPolicySettings 配置模型，含规则的 service_tier × action × scope
× 模型白名单 × fallback action 维度。
- SettingService.Get/SetOpenAIFastPolicySettings；缺失时返回内置默认策略
（所有模型的 priority 走 filter，whitelist 为空，fallback=pass）。设计
依据：service_tier=fast 是用户级开关，与 model 字段正交，默认锁定特定
model slug 会留下"用 gpt-4 + fast 透传 priority 上游"的绕过路径。JSON
解析失败不再静默 fallback，slog.Warn 记录脏数据，便于运维定位。
- service_tier 归一化（trim + ToLower + fast→priority + 白名单 priority/flex）
与策略评估（evaluateOpenAIFastPolicy）作为唯一真实来源，HTTP / WS 共用。
抽出纯函数 evaluateOpenAIFastPolicyWithSettings，配合 ctx-bound settings
快照（withOpenAIFastPolicyContext / openAIFastPolicySettingsFromContext），
WS 长会话入口预取一次后所有帧复用，避免每帧打到 settingService。
HTTP 入口（4 个）
- Chat Completions、Anthropic 兼容（Messages，含 BetaFastMode→priority 二次
命中）、原生 Responses、Passthrough Responses 全部接入
applyOpenAIFastPolicyToBody，filter 走 sjson 顶层删除 service_tier，block
返回 403 forbidden_error JSON。
- 4 入口统一使用 upstream 视角的 model（GetMappedModel +
normalizeOpenAIModelForUpstream + Codex OAuth normalize 后的 slug），
避免 chat/messages/native /responses/passthrough 因为 model 维度不同
造成 whitelist 命中差异。
- 在 pass 路径也把客户端 "fast" 别名归一化为 "priority" 写回 body，
否则 native /responses 与 passthrough 入口会把 "fast" 原样透传给上游
导致 400/拒绝（chat-completions 入口的 normalizeResponsesBodyServiceTier
此前已具备同等行为）。
WebSocket 入口
- 新增 applyOpenAIFastPolicyToWSResponseCreate：严格匹配
type="response.create"，仅处理顶层 service_tier；filter 用 sjson 删字段，
block 返回 typed *OpenAIFastBlockedError。
- ingress 路径在 parseClientPayload 内调用，block 命中先 Write Realtime
风格 error event 再返回 OpenAIWSClientCloseError(StatusPolicyViolation
=1008)，依赖底层 WebSocket Conn.Write 的同步 flush 保证 error 先于
close。
- passthrough 路径在 RunEntry 前对 firstClientMessage 应用策略，并通过
openAIWSPolicyEnforcingFrameConn 包装 ReadFrame 对每个 client→upstream
帧执行策略；后续帧无 model 字段时回退到 capturedSessionModel。
filter 闭包内同时侦测 session.update / session.created 帧的 session.model
字段刷新 capturedSessionModel，封堵"首帧 model=gpt-4o（pass）→
session.update 改为 gpt-5.5 → 不带 model 的 response.create fallback
到 gpt-4o"的 mid-session 绕过路径。
- passthrough billing：requestServiceTier 在策略 filter 之后再从
firstClientMessage 提取，filter 命中时 OpenAIForwardResult.ServiceTier
上报 nil（default tier），与 HTTP 入口（reqBody 来自 post-filter map）
/ WS ingress（payload 来自 post-filter bytes）的语义一致。
- 错误事件 schema：{event_id: "evt_<32hex>", type: "error",
error: {type: "forbidden_error", code: "policy_violation", message}}，
与 OpenAI codex 客户端 error event 解析兼容。
Admin / Frontend
- dto.SystemSettings / UpdateSettingsRequest 新增
openai_fast_policy_settings 字段（omitempty），bulk GET/PUT 接入。
- Settings 页 Gateway 页签新增 Fast/Flex Policy 表单卡片：
service_tier × action × scope × 模型白名单 × fallback action 全字段配置。
- 前端守门：openaiFastPolicyLoaded 标志仅在 GET 真带回字段时才允许回写，
避免 rollout/错误把默认规则覆盖成空；saveSettings 回写循环 skip 该字段，
由专用刷新逻辑处理；仅 action=block 时发送 error_message，匹配后端
omitempty 行为。
测试
- HTTP 路径：openai_fast_policy_test.go 覆盖默认配置（whitelist=[]，所有
模型 priority filter）/ block 自定义错误 / scope 区分 / filter 删字段 /
block 不改 body / block 短路上游 / Anthropic BetaFastMode 触发 OpenAI
fast policy 等场景。
- WebSocket 路径：openai_fast_policy_ws_test.go 覆盖
helper 单元（filter / fast→priority 归一化 / flex 透传 / block typed
error / 无 service_tier 字节不变 / 非 response.create 帧不动 / 空 type
帧不动 / event_id+code 字段断言 / 非字符串 service_tier 容错）+
pass 路径 fast 别名归一化回归 +
ingress 端到端（filter 后上游不含 service_tier / block 后客户端先收
error event 再收 close 1008 且上游 0 写）+
passthrough capturedSessionModel fallback 用例（whitelist 策略下首帧
建立、缺 model 命中 fallback、缺少 fallback 时的 leak 文档化）+
passthrough session.update / session.created 旋转 capturedSessionModel
的 mid-session 绕过回归 +
passthrough billing post-filter ServiceTier 与 idempotent filter 回归。
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
DaydreamCoding
@@ -2366,6 +2366,15 @@ func (s *OpenAIGatewayService) ProxyResponsesWebSocketFromClient(
|
||||
return errors.New("token is empty")
|
||||
}
|
||||
|
||||
// 预取一次 OpenAI Fast Policy settings，绑定到 ctx，让该 WS session
|
||||
// 内所有帧的 evaluateOpenAIFastPolicy 调用复用同一份快照，避免每帧
|
||||
// 进入 DB / settingRepo。Trade-off 见 withOpenAIFastPolicyContext 注释。
|
||||
if s.settingService != nil {
|
||||
if settings, err := s.settingService.GetOpenAIFastPolicySettings(ctx); err == nil && settings != nil {
|
||||
ctx = withOpenAIFastPolicyContext(ctx, settings)
|
||||
}
|
||||
}
|
||||
|
||||
wsDecision := s.getOpenAIWSProtocolResolver().Resolve(account)
|
||||
modeRouterV2Enabled := s != nil && s.cfg != nil && s.cfg.Gateway.OpenAIWS.ModeRouterV2Enabled
|
||||
ingressMode := OpenAIWSIngressModeCtxPool
|
||||
@@ -2524,6 +2533,44 @@ func (s *OpenAIGatewayService) ProxyResponsesWebSocketFromClient(
|
||||
normalized = next
|
||||
}
|
||||
|
||||
// Apply OpenAI Fast Policy on the response.create frame using the same
|
||||
// evaluator/normalize/scope rules as the HTTP entrypoints. This is the
|
||||
// single integration point for all WS ingress turns (first + follow-up
|
||||
// frames flow through here).
|
||||
//
|
||||
// Model fallback: parseClientPayload above rejects any frame whose
|
||||
// "model" field is missing (line ~2493-2500), so by the time we
|
||||
// reach this point upstreamModel is always derived from a non-empty
|
||||
// per-frame model. The capturedSessionModel fallback used in the
|
||||
// passthrough adapter is therefore not needed in this path.
|
||||
policyApplied, blocked, policyErr := s.applyOpenAIFastPolicyToWSResponseCreate(ctx, account, upstreamModel, normalized)
|
||||
if policyErr != nil {
|
||||
return openAIWSClientPayload{}, NewOpenAIWSClientCloseError(coderws.StatusPolicyViolation, "invalid websocket request payload", policyErr)
|
||||
}
|
||||
if blocked != nil {
|
||||
// Send a Realtime-style error event to the client first, then
|
||||
// signal the handler to close the connection with PolicyViolation.
|
||||
// We intentionally do NOT forward this frame upstream.
|
||||
//
|
||||
// coder/websocket@v1.8.14 Conn.Write is synchronous and flushes
|
||||
// the underlying bufio writer before returning (write.go:42 →
|
||||
// 307-311), and the subsequent close handshake re-acquires the
|
||||
// same writeFrameMu, so the error event is guaranteed to reach
|
||||
// the kernel send buffer before any close frame is queued.
|
||||
eventBytes := buildOpenAIFastPolicyBlockedWSEvent(blocked)
|
||||
if eventBytes != nil {
|
||||
writeCtx, cancel := context.WithTimeout(ctx, s.openAIWSWriteTimeout())
|
||||
_ = clientConn.Write(writeCtx, coderws.MessageText, eventBytes)
|
||||
cancel()
|
||||
}
|
||||
return openAIWSClientPayload{}, NewOpenAIWSClientCloseError(
|
||||
coderws.StatusPolicyViolation,
|
||||
blocked.Message,
|
||||
blocked,
|
||||
)
|
||||
}
|
||||
normalized = policyApplied
|
||||
|
||||
return openAIWSClientPayload{
|
||||
payloadRaw: normalized,
|
||||
rawForHash: trimmed,
|
||||
|
||||
Reference in New Issue
Block a user