diff --git a/.github/workflows/release.yaml b/.github/workflows/release-stable.yaml similarity index 98% rename from .github/workflows/release.yaml rename to .github/workflows/release-stable.yaml index 4043e4a5dd..201c1cabe3 100644 --- a/.github/workflows/release.yaml +++ b/.github/workflows/release-stable.yaml @@ -4,7 +4,7 @@ on: push: # run only against tags tags: - - '*' + - 'v*' permissions: contents: write diff --git a/README.md b/README.md index 9e15436f45..068e98e18b 100644 --- a/README.md +++ b/README.md @@ -12,7 +12,101 @@ This is the Plus version of [CLIProxyAPI](https://github.com/router-for-me/CLIPr All third-party provider support is maintained by community contributors; CLIProxyAPI does not provide technical support. Please contact the corresponding community maintainer if you need assistance. -The Plus release stays in lockstep with the mainline features. +[![z.ai](https://assets.router-for.me/english-5-0.jpg)](https://z.ai/subscribe?ic=8JVLJQFSKB) + +This project is sponsored by Z.ai, supporting us with their GLM CODING PLAN. + +GLM CODING PLAN is a subscription service designed for AI coding, starting at just $10/month. It provides access to their flagship GLM-4.7 & （GLM-5 Only Available for Pro Users）model across 10+ popular AI coding tools (Claude Code, Cline, Roo Code, etc.), offering developers top-tier, fast, and stable coding experiences. + +Get 10% OFF GLM CODING PLAN：https://z.ai/subscribe?ic=8JVLJQFSKB + +--- + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
PackyCodeThanks to PackyCode for sponsoring this project! PackyCode is a reliable and efficient API relay service provider, offering relay services for Claude Code, Codex, Gemini, and more. PackyCode provides special discounts for our software users: register using this link and enter the "cliproxyapi" promo code during recharge to get 10% off.
AICodeMirrorThanks to AICodeMirror for sponsoring this project! AICodeMirror provides official high-stability relay services for Claude Code / Codex / Gemini CLI, with enterprise-grade concurrency, fast invoicing, and 24/7 dedicated technical support. Claude Code / Codex / Gemini official channels at 38% / 2% / 9% of original price, with extra discounts on top-ups! AICodeMirror offers special benefits for CLIProxyAPI users: register via this link to enjoy 20% off your first top-up, and enterprise customers can get up to 25% off!
BmoPlusHuge thanks to BmoPlus for sponsoring this project! BmoPlus is a highly reliable AI account provider built strictly for heavy AI users and developers. They offer rock-solid, ready-to-use accounts and official top-up services for ChatGPT Plus / ChatGPT Pro (Full Warranty) / Claude Pro / Super Grok / Gemini Pro. By registering and ordering through BmoPlus - Premium AI Accounts & Top-ups, users can unlock the mind-blowing rate of 10% of the official GPT subscription price (90% OFF)!
LingtrueAPIThanks to LingtrueAPI for its sponsorship of this project! LingtrueAPI is a global large - model API intermediary service platform that provides API calling services for various top - notch models such as Claude Code, Codex, and Gemini. It is committed to enabling users to connect to global AI capabilities at low cost and with high stability. LingtrueAPI offers special discounts to users of this software: register using this link, and enter the promo code "LingtrueAPI" when making the first recharge to enjoy a 10% discount.
PoixeAIThanks to Poixe AI for sponsoring this project! Poixe AI provides reliable LLM API services. You can leverage the platform's API endpoints to seamlessly build AI-powered products. Additionally, you can become a vendor by providing AI API resources to the platform and earn revenue. Register through the exclusive CLIProxyAPI referral link and receive a bonus of $5 USD on your first top-up.
VisionCoderThanks to VisionCoder for supporting this project. VisionCoder Developer Platform is a reliable and efficient API relay service provider, offering access to mainstream AI models such as Claude Code, Codex, and Gemini. It helps developers and teams integrate AI capabilities more easily and improve productivity. +

+VisionCoder is also offering our users a limited-time Token Plan promotion: buy 1 month and get 1 month free.
+ +## Overview + +- OpenAI/Gemini/Claude compatible API endpoints for CLI models +- OpenAI Codex support (GPT models) via OAuth login +- Claude Code support via OAuth login +- Amp CLI and IDE extensions support with provider routing +- Streaming and non-streaming responses +- Function calling/tools support +- Multimodal input support (text and images) +- Multiple accounts with round-robin load balancing (Gemini, OpenAI, Claude) +- Simple CLI authentication flows (Gemini, OpenAI, Claude) +- Generative Language API Key support +- AI Studio Build multi-account load balancing +- Gemini CLI multi-account load balancing +- Claude Code multi-account load balancing +- OpenAI Codex multi-account load balancing +- OpenAI-compatible upstream providers via config (e.g., OpenRouter) +- Reusable Go SDK for embedding the proxy (see `docs/sdk-usage.md`) + +## Getting Started + +CLIProxyAPI Guides: [https://help.router-for.me/](https://help.router-for.me/) + +## Management API + +see [MANAGEMENT_API.md](https://help.router-for.me/management/api) + +## Amp CLI Support + +CLIProxyAPI includes integrated support for [Amp CLI](https://ampcode.com) and Amp IDE extensions, enabling you to use your Google/ChatGPT/Claude OAuth subscriptions with Amp's coding tools: + +- Provider route aliases for Amp's API patterns (`/api/provider/{provider}/v1...`) +- Management proxy for OAuth authentication and account features +- Smart model fallback with automatic routing +- **Model mapping** to route unavailable models to alternatives (e.g., `claude-opus-4.5` → `claude-sonnet-4`) +- Security-first design with localhost-only management endpoints + +When you need the request/response shape of a specific backend family, use the provider-specific paths instead of the merged `/v1/...` endpoints: + +- Use `/api/provider/{provider}/v1/messages` for messages-style backends. +- Use `/api/provider/{provider}/v1beta/models/...` for model-scoped generate endpoints. +- Use `/api/provider/{provider}/v1/chat/completions` for chat-completions backends. + +These routes help you select the protocol surface, but they do not by themselves guarantee a unique inference executor when the same client-visible model name is reused across multiple backends. Inference routing is still resolved from the request model/alias. For strict backend pinning, use unique aliases, prefixes, or otherwise avoid overlapping client-visible model names. + +**→ [Complete Amp CLI Integration Guide](https://help.router-for.me/agent-client/amp-cli.html)** + +## SDK Docs + +- Usage: [docs/sdk-usage.md](docs/sdk-usage.md) +- Advanced (executors & translators): [docs/sdk-advanced.md](docs/sdk-advanced.md) +- Access: [docs/sdk-access.md](docs/sdk-access.md) +- Watcher: [docs/sdk-watcher.md](docs/sdk-watcher.md) +- Custom Provider Example: `examples/custom-provider` ## Contributing diff --git a/README_CN.md b/README_CN.md index 5c0bce578d..94aa557a43 100644 --- a/README_CN.md +++ b/README_CN.md @@ -7,7 +7,108 @@ 这是 [CLIProxyAPI](https://github.com/router-for-me/CLIProxyAPI) 的 Plus 版本，在原有基础上增加了第三方供应商的支持。 -所有的第三方供应商支持都由第三方社区维护者提供，CLIProxyAPI 不提供技术支持。如需取得支持，请与对应的社区维护者联系。 +一个为 CLI 提供 OpenAI/Gemini/Claude/Codex 兼容 API 接口的代理服务器。 + +现已支持通过 OAuth 登录接入 OpenAI Codex（GPT 系列）和 Claude Code。 + +您可以使用本地或多账户的CLI方式，通过任何与 OpenAI（包括Responses）/Gemini/Claude 兼容的客户端和SDK进行访问。 + +## 赞助商 + +[![bigmodel.cn](https://assets.router-for.me/chinese-5-0.jpg)](https://www.bigmodel.cn/claude-code?ic=RRVJPB5SII) + +本项目由 Z智谱 提供赞助, 他们通过 GLM CODING PLAN 对本项目提供技术支持。 + +GLM CODING PLAN 是专为AI编码打造的订阅套餐，每月最低仅需20元，即可在十余款主流AI编码工具如 Claude Code、Cline、Roo Code 中畅享智谱旗舰模型GLM-4.7（受限于算力，目前仅限Pro用户开放），为开发者提供顶尖的编码体验。 + +智谱AI为本产品提供了特别优惠，使用以下链接购买可以享受九折优惠：https://www.bigmodel.cn/claude-code?ic=RRVJPB5SII + +--- + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
PackyCode感谢 PackyCode 对本项目的赞助！PackyCode 是一家可靠高效的 API 中转服务商，提供 Claude Code、Codex、Gemini 等多种服务的中转。PackyCode 为本软件用户提供了特别优惠：使用此链接注册，并在充值时输入 "cliproxyapi" 优惠码即可享受九折优惠。
AICodeMirror感谢 AICodeMirror 赞助了本项目！AICodeMirror 提供 Claude Code / Codex / Gemini CLI 官方高稳定中转服务，支持企业级高并发、极速开票、7×24 专属技术支持。 Claude Code / Codex / Gemini 官方渠道低至 3.8 / 0.2 / 0.9 折，充值更有折上折！AICodeMirror 为 CLIProxyAPI 的用户提供了特别福利，通过此链接注册的用户，可享受首充8折，企业客户最高可享 7.5 折！
BmoPlus感谢 BmoPlus 赞助了本项目！BmoPlus 是一家专为AI订阅重度用户打造的可靠 AI 账号代充服务商，提供稳定的 ChatGPT Plus / ChatGPT Pro(全程质保) / Claude Pro / Super Grok / Gemini Pro 的官方代充&成品账号。 通过BmoPlus AI成品号专卖/代充注册下单的用户，可享GPT 官网订阅一折 的震撼价格！
LingtrueAPI感谢 LingtrueAPI 对本项目的赞助！LingtrueAPI 是一家全球大模型API中转服务平台，提供Claude Code、Codex、Gemini 等多种顶级模型API调用服务，致力于让用户以低成本、高稳定性链接全球AI能力。LingtrueAPI为本软件用户提供了特别优惠：使用此链接注册，并在首次充值时输入 "LingtrueAPI" 优惠码即可享受9折优惠。
PoixeAI感谢 Poixe AI 对本项目的赞助！Poixe AI 提供可靠的 AI 模型接口服务，您可以使用平台提供的 LLM API 接口轻松构建 AI 产品，同时也可以成为供应商，为平台提供大模型资源以赚取收益。通过 CLIProxyAPI 专属链接注册，充值额外赠送 $5 美金
VisionCoder感谢 VisionCoder 对本项目的支持。VisionCoder 开发平台 是一个可靠高效的 API 中继服务提供商，提供 Claude Code、Codex、Gemini 等主流 AI 模型，帮助开发者和团队更轻松地集成 AI 功能，提升工作效率。 +

+VisionCoder 还为我们的用户提供 Token Plan 限时活动：购买 1 个月，赠送 1 个月。
+ + +## 功能特性 + +- 为 CLI 模型提供 OpenAI/Gemini/Claude/Codex 兼容的 API 端点 +- 新增 OpenAI Codex（GPT 系列）支持（OAuth 登录） +- 新增 Claude Code 支持（OAuth 登录） +- 支持流式与非流式响应 +- 函数调用/工具支持 +- 多模态输入（文本、图片） +- 多账户支持与轮询负载均衡（Gemini、OpenAI、Claude） +- 简单的 CLI 身份验证流程（Gemini、OpenAI、Claude） +- 支持 Gemini AIStudio API 密钥 +- 支持 AI Studio Build 多账户轮询 +- 支持 Gemini CLI 多账户轮询 +- 支持 Claude Code 多账户轮询 +- 支持 OpenAI Codex 多账户轮询 +- 通过配置接入上游 OpenAI 兼容提供商（例如 OpenRouter） +- 可复用的 Go SDK（见 `docs/sdk-usage_CN.md`） + +## 新手入门 + +CLIProxyAPI 用户手册： [https://help.router-for.me/](https://help.router-for.me/cn/) + +## 管理 API 文档 + +请参见 [MANAGEMENT_API_CN.md](https://help.router-for.me/cn/management/api) + +## Amp CLI 支持 + +CLIProxyAPI 已内置对 [Amp CLI](https://ampcode.com) 和 Amp IDE 扩展的支持，可让你使用自己的 Google/ChatGPT/Claude OAuth 订阅来配合 Amp 编码工具： + +- 提供商路由别名，兼容 Amp 的 API 路径模式（`/api/provider/{provider}/v1...`） +- 管理代理，处理 OAuth 认证和账号功能 +- 智能模型回退与自动路由 +- 以安全为先的设计，管理端点仅限 localhost + +当你需要某一类后端的请求/响应协议形态时，优先使用 provider-specific 路径，而不是合并后的 `/v1/...` 端点： + +- 对于 messages 风格的后端，使用 `/api/provider/{provider}/v1/messages`。 +- 对于按模型路径暴露生成接口的后端，使用 `/api/provider/{provider}/v1beta/models/...`。 +- 对于 chat-completions 风格的后端，使用 `/api/provider/{provider}/v1/chat/completions`。 + +这些路径有助于选择协议表面，但当多个后端复用同一个客户端可见模型名时，它们本身并不能保证唯一的推理执行器。实际的推理路由仍然根据请求里的 model/alias 解析。若要严格钉住某个后端，请使用唯一 alias、前缀，或避免让多个后端暴露相同的客户端模型名。 + +**→ [Amp CLI 完整集成指南](https://help.router-for.me/cn/agent-client/amp-cli.html)** + +## SDK 文档 + +- 使用文档：[docs/sdk-usage_CN.md](docs/sdk-usage_CN.md) +- 高级（执行器与翻译器）：[docs/sdk-advanced_CN.md](docs/sdk-advanced_CN.md) +- 认证: [docs/sdk-access_CN.md](docs/sdk-access_CN.md) +- 凭据加载/更新: [docs/sdk-watcher_CN.md](docs/sdk-watcher_CN.md) +- 自定义 Provider 示例：`examples/custom-provider` ## 贡献 diff --git a/assets/visioncoder.png b/assets/visioncoder.png new file mode 100644 index 0000000000..24b1760ce5 Binary files /dev/null and b/assets/visioncoder.png differ diff --git a/internal/api/handlers/management/config_auth_index.go b/internal/api/handlers/management/config_auth_index.go new file mode 100644 index 0000000000..ed0b3ec42d --- /dev/null +++ b/internal/api/handlers/management/config_auth_index.go @@ -0,0 +1,241 @@ +package management + +import ( + "fmt" + "strings" + + "github.com/router-for-me/CLIProxyAPI/v6/internal/config" + "github.com/router-for-me/CLIProxyAPI/v6/internal/watcher/synthesizer" +) + +type geminiKeyWithAuthIndex struct { + config.GeminiKey + AuthIndex string `json:"auth-index,omitempty"` +} + +type claudeKeyWithAuthIndex struct { + config.ClaudeKey + AuthIndex string `json:"auth-index,omitempty"` +} + +type codexKeyWithAuthIndex struct { + config.CodexKey + AuthIndex string `json:"auth-index,omitempty"` +} + +type vertexCompatKeyWithAuthIndex struct { + config.VertexCompatKey + AuthIndex string `json:"auth-index,omitempty"` +} + +type openAICompatibilityAPIKeyWithAuthIndex struct { + config.OpenAICompatibilityAPIKey + AuthIndex string `json:"auth-index,omitempty"` +} + +type openAICompatibilityWithAuthIndex struct { + Name string `json:"name"` + Priority int `json:"priority,omitempty"` + Prefix string `json:"prefix,omitempty"` + BaseURL string `json:"base-url"` + APIKeyEntries []openAICompatibilityAPIKeyWithAuthIndex `json:"api-key-entries,omitempty"` + Models []config.OpenAICompatibilityModel `json:"models,omitempty"` + Headers map[string]string `json:"headers,omitempty"` + AuthIndex string `json:"auth-index,omitempty"` +} + +func (h *Handler) liveAuthIndexByID() map[string]string { + out := map[string]string{} + if h == nil { + return out + } + h.mu.Lock() + manager := h.authManager + h.mu.Unlock() + if manager == nil { + return out + } + // authManager.List() returns clones, so EnsureIndex only affects these copies. + for _, auth := range manager.List() { + if auth == nil { + continue + } + id := strings.TrimSpace(auth.ID) + if id == "" { + continue + } + idx := strings.TrimSpace(auth.Index) + if idx == "" { + idx = auth.EnsureIndex() + } + if idx == "" { + continue + } + out[id] = idx + } + return out +} + +func (h *Handler) geminiKeysWithAuthIndex() []geminiKeyWithAuthIndex { + if h == nil { + return nil + } + liveIndexByID := h.liveAuthIndexByID() + + h.mu.Lock() + defer h.mu.Unlock() + if h.cfg == nil { + return nil + } + + idGen := synthesizer.NewStableIDGenerator() + out := make([]geminiKeyWithAuthIndex, len(h.cfg.GeminiKey)) + for i := range h.cfg.GeminiKey { + entry := h.cfg.GeminiKey[i] + authIndex := "" + if key := strings.TrimSpace(entry.APIKey); key != "" { + id, _ := idGen.Next("gemini:apikey", key, entry.BaseURL) + authIndex = liveIndexByID[id] + } + out[i] = geminiKeyWithAuthIndex{ + GeminiKey: entry, + AuthIndex: authIndex, + } + } + return out +} + +func (h *Handler) claudeKeysWithAuthIndex() []claudeKeyWithAuthIndex { + if h == nil { + return nil + } + liveIndexByID := h.liveAuthIndexByID() + + h.mu.Lock() + defer h.mu.Unlock() + if h.cfg == nil { + return nil + } + + idGen := synthesizer.NewStableIDGenerator() + out := make([]claudeKeyWithAuthIndex, len(h.cfg.ClaudeKey)) + for i := range h.cfg.ClaudeKey { + entry := h.cfg.ClaudeKey[i] + authIndex := "" + if key := strings.TrimSpace(entry.APIKey); key != "" { + id, _ := idGen.Next("claude:apikey", key, entry.BaseURL) + authIndex = liveIndexByID[id] + } + out[i] = claudeKeyWithAuthIndex{ + ClaudeKey: entry, + AuthIndex: authIndex, + } + } + return out +} + +func (h *Handler) codexKeysWithAuthIndex() []codexKeyWithAuthIndex { + if h == nil { + return nil + } + liveIndexByID := h.liveAuthIndexByID() + + h.mu.Lock() + defer h.mu.Unlock() + if h.cfg == nil { + return nil + } + + idGen := synthesizer.NewStableIDGenerator() + out := make([]codexKeyWithAuthIndex, len(h.cfg.CodexKey)) + for i := range h.cfg.CodexKey { + entry := h.cfg.CodexKey[i] + authIndex := "" + if key := strings.TrimSpace(entry.APIKey); key != "" { + id, _ := idGen.Next("codex:apikey", key, entry.BaseURL) + authIndex = liveIndexByID[id] + } + out[i] = codexKeyWithAuthIndex{ + CodexKey: entry, + AuthIndex: authIndex, + } + } + return out +} + +func (h *Handler) vertexCompatKeysWithAuthIndex() []vertexCompatKeyWithAuthIndex { + if h == nil { + return nil + } + liveIndexByID := h.liveAuthIndexByID() + + h.mu.Lock() + defer h.mu.Unlock() + if h.cfg == nil { + return nil + } + + idGen := synthesizer.NewStableIDGenerator() + out := make([]vertexCompatKeyWithAuthIndex, len(h.cfg.VertexCompatAPIKey)) + for i := range h.cfg.VertexCompatAPIKey { + entry := h.cfg.VertexCompatAPIKey[i] + id, _ := idGen.Next("vertex:apikey", entry.APIKey, entry.BaseURL, entry.ProxyURL) + authIndex := liveIndexByID[id] + out[i] = vertexCompatKeyWithAuthIndex{ + VertexCompatKey: entry, + AuthIndex: authIndex, + } + } + return out +} + +func (h *Handler) openAICompatibilityWithAuthIndex() []openAICompatibilityWithAuthIndex { + if h == nil { + return nil + } + liveIndexByID := h.liveAuthIndexByID() + + h.mu.Lock() + defer h.mu.Unlock() + if h.cfg == nil { + return nil + } + + normalized := normalizedOpenAICompatibilityEntries(h.cfg.OpenAICompatibility) + out := make([]openAICompatibilityWithAuthIndex, len(normalized)) + idGen := synthesizer.NewStableIDGenerator() + for i := range normalized { + entry := normalized[i] + providerName := strings.ToLower(strings.TrimSpace(entry.Name)) + if providerName == "" { + providerName = "openai-compatibility" + } + idKind := fmt.Sprintf("openai-compatibility:%s", providerName) + + response := openAICompatibilityWithAuthIndex{ + Name: entry.Name, + Priority: entry.Priority, + Prefix: entry.Prefix, + BaseURL: entry.BaseURL, + Models: entry.Models, + Headers: entry.Headers, + AuthIndex: "", + } + if len(entry.APIKeyEntries) == 0 { + id, _ := idGen.Next(idKind, entry.BaseURL) + response.AuthIndex = liveIndexByID[id] + } else { + response.APIKeyEntries = make([]openAICompatibilityAPIKeyWithAuthIndex, len(entry.APIKeyEntries)) + for j := range entry.APIKeyEntries { + apiKeyEntry := entry.APIKeyEntries[j] + id, _ := idGen.Next(idKind, apiKeyEntry.APIKey, entry.BaseURL, apiKeyEntry.ProxyURL) + response.APIKeyEntries[j] = openAICompatibilityAPIKeyWithAuthIndex{ + OpenAICompatibilityAPIKey: apiKeyEntry, + AuthIndex: liveIndexByID[id], + } + } + } + out[i] = response + } + return out +} diff --git a/internal/api/handlers/management/config_lists.go b/internal/api/handlers/management/config_lists.go index a1a95b2cd2..b3eaaa800d 100644 --- a/internal/api/handlers/management/config_lists.go +++ b/internal/api/handlers/management/config_lists.go @@ -120,7 +120,7 @@ func (h *Handler) DeleteAPIKeys(c *gin.Context) { // gemini-api-key: []GeminiKey func (h *Handler) GetGeminiKeys(c *gin.Context) { - c.JSON(200, gin.H{"gemini-api-key": h.cfg.GeminiKey}) + c.JSON(200, gin.H{"gemini-api-key": h.geminiKeysWithAuthIndex()}) } func (h *Handler) PutGeminiKeys(c *gin.Context) { data, err := c.GetRawData() @@ -139,9 +139,11 @@ func (h *Handler) PutGeminiKeys(c *gin.Context) { } arr = obj.Items } + h.mu.Lock() + defer h.mu.Unlock() h.cfg.GeminiKey = append([]config.GeminiKey(nil), arr...) h.cfg.SanitizeGeminiKeys() - h.persist(c) + h.persistLocked(c) } func (h *Handler) PatchGeminiKey(c *gin.Context) { type geminiKeyPatch struct { @@ -161,6 +163,9 @@ func (h *Handler) PatchGeminiKey(c *gin.Context) { c.JSON(400, gin.H{"error": "invalid body"}) return } + + h.mu.Lock() + defer h.mu.Unlock() targetIndex := -1 if body.Index != nil && *body.Index >= 0 && *body.Index < len(h.cfg.GeminiKey) { targetIndex = *body.Index @@ -187,7 +192,7 @@ func (h *Handler) PatchGeminiKey(c *gin.Context) { if trimmed == "" { h.cfg.GeminiKey = append(h.cfg.GeminiKey[:targetIndex], h.cfg.GeminiKey[targetIndex+1:]...) h.cfg.SanitizeGeminiKeys() - h.persist(c) + h.persistLocked(c) return } entry.APIKey = trimmed @@ -209,10 +214,12 @@ func (h *Handler) PatchGeminiKey(c *gin.Context) { } h.cfg.GeminiKey[targetIndex] = entry h.cfg.SanitizeGeminiKeys() - h.persist(c) + h.persistLocked(c) } func (h *Handler) DeleteGeminiKey(c *gin.Context) { + h.mu.Lock() + defer h.mu.Unlock() if val := strings.TrimSpace(c.Query("api-key")); val != "" { if baseRaw, okBase := c.GetQuery("base-url"); okBase { base := strings.TrimSpace(baseRaw) @@ -226,7 +233,7 @@ func (h *Handler) DeleteGeminiKey(c *gin.Context) { if len(out) != len(h.cfg.GeminiKey) { h.cfg.GeminiKey = out h.cfg.SanitizeGeminiKeys() - h.persist(c) + h.persistLocked(c) } else { c.JSON(404, gin.H{"error": "item not found"}) } @@ -253,7 +260,7 @@ func (h *Handler) DeleteGeminiKey(c *gin.Context) { } h.cfg.GeminiKey = append(h.cfg.GeminiKey[:matchIndex], h.cfg.GeminiKey[matchIndex+1:]...) h.cfg.SanitizeGeminiKeys() - h.persist(c) + h.persistLocked(c) return } if idxStr := c.Query("index"); idxStr != "" { @@ -261,7 +268,7 @@ func (h *Handler) DeleteGeminiKey(c *gin.Context) { if _, err := fmt.Sscanf(idxStr, "%d", &idx); err == nil && idx >= 0 && idx < len(h.cfg.GeminiKey) { h.cfg.GeminiKey = append(h.cfg.GeminiKey[:idx], h.cfg.GeminiKey[idx+1:]...) h.cfg.SanitizeGeminiKeys() - h.persist(c) + h.persistLocked(c) return } } @@ -270,7 +277,7 @@ func (h *Handler) DeleteGeminiKey(c *gin.Context) { // claude-api-key: []ClaudeKey func (h *Handler) GetClaudeKeys(c *gin.Context) { - c.JSON(200, gin.H{"claude-api-key": h.cfg.ClaudeKey}) + c.JSON(200, gin.H{"claude-api-key": h.claudeKeysWithAuthIndex()}) } func (h *Handler) PutClaudeKeys(c *gin.Context) { data, err := c.GetRawData() @@ -292,9 +299,11 @@ func (h *Handler) PutClaudeKeys(c *gin.Context) { for i := range arr { normalizeClaudeKey(&arr[i]) } + h.mu.Lock() + defer h.mu.Unlock() h.cfg.ClaudeKey = arr h.cfg.SanitizeClaudeKeys() - h.persist(c) + h.persistLocked(c) } func (h *Handler) PatchClaudeKey(c *gin.Context) { type claudeKeyPatch struct { @@ -315,6 +324,9 @@ func (h *Handler) PatchClaudeKey(c *gin.Context) { c.JSON(400, gin.H{"error": "invalid body"}) return } + + h.mu.Lock() + defer h.mu.Unlock() targetIndex := -1 if body.Index != nil && *body.Index >= 0 && *body.Index < len(h.cfg.ClaudeKey) { targetIndex = *body.Index @@ -358,10 +370,12 @@ func (h *Handler) PatchClaudeKey(c *gin.Context) { normalizeClaudeKey(&entry) h.cfg.ClaudeKey[targetIndex] = entry h.cfg.SanitizeClaudeKeys() - h.persist(c) + h.persistLocked(c) } func (h *Handler) DeleteClaudeKey(c *gin.Context) { + h.mu.Lock() + defer h.mu.Unlock() if val := strings.TrimSpace(c.Query("api-key")); val != "" { if baseRaw, okBase := c.GetQuery("base-url"); okBase { base := strings.TrimSpace(baseRaw) @@ -374,7 +388,7 @@ func (h *Handler) DeleteClaudeKey(c *gin.Context) { } h.cfg.ClaudeKey = out h.cfg.SanitizeClaudeKeys() - h.persist(c) + h.persistLocked(c) return } @@ -396,7 +410,7 @@ func (h *Handler) DeleteClaudeKey(c *gin.Context) { h.cfg.ClaudeKey = append(h.cfg.ClaudeKey[:matchIndex], h.cfg.ClaudeKey[matchIndex+1:]...) } h.cfg.SanitizeClaudeKeys() - h.persist(c) + h.persistLocked(c) return } if idxStr := c.Query("index"); idxStr != "" { @@ -405,7 +419,7 @@ func (h *Handler) DeleteClaudeKey(c *gin.Context) { if err == nil && idx >= 0 && idx < len(h.cfg.ClaudeKey) { h.cfg.ClaudeKey = append(h.cfg.ClaudeKey[:idx], h.cfg.ClaudeKey[idx+1:]...) h.cfg.SanitizeClaudeKeys() - h.persist(c) + h.persistLocked(c) return } } @@ -414,7 +428,7 @@ func (h *Handler) DeleteClaudeKey(c *gin.Context) { // openai-compatibility: []OpenAICompatibility func (h *Handler) GetOpenAICompat(c *gin.Context) { - c.JSON(200, gin.H{"openai-compatibility": normalizedOpenAICompatibilityEntries(h.cfg.OpenAICompatibility)}) + c.JSON(200, gin.H{"openai-compatibility": h.openAICompatibilityWithAuthIndex()}) } func (h *Handler) PutOpenAICompat(c *gin.Context) { data, err := c.GetRawData() @@ -440,9 +454,11 @@ func (h *Handler) PutOpenAICompat(c *gin.Context) { filtered = append(filtered, arr[i]) } } + h.mu.Lock() + defer h.mu.Unlock() h.cfg.OpenAICompatibility = filtered h.cfg.SanitizeOpenAICompatibility() - h.persist(c) + h.persistLocked(c) } func (h *Handler) PatchOpenAICompat(c *gin.Context) { type openAICompatPatch struct { @@ -462,6 +478,9 @@ func (h *Handler) PatchOpenAICompat(c *gin.Context) { c.JSON(400, gin.H{"error": "invalid body"}) return } + + h.mu.Lock() + defer h.mu.Unlock() targetIndex := -1 if body.Index != nil && *body.Index >= 0 && *body.Index < len(h.cfg.OpenAICompatibility) { targetIndex = *body.Index @@ -492,7 +511,7 @@ func (h *Handler) PatchOpenAICompat(c *gin.Context) { if trimmed == "" { h.cfg.OpenAICompatibility = append(h.cfg.OpenAICompatibility[:targetIndex], h.cfg.OpenAICompatibility[targetIndex+1:]...) h.cfg.SanitizeOpenAICompatibility() - h.persist(c) + h.persistLocked(c) return } entry.BaseURL = trimmed @@ -509,10 +528,12 @@ func (h *Handler) PatchOpenAICompat(c *gin.Context) { normalizeOpenAICompatibilityEntry(&entry) h.cfg.OpenAICompatibility[targetIndex] = entry h.cfg.SanitizeOpenAICompatibility() - h.persist(c) + h.persistLocked(c) } func (h *Handler) DeleteOpenAICompat(c *gin.Context) { + h.mu.Lock() + defer h.mu.Unlock() if name := c.Query("name"); name != "" { out := make([]config.OpenAICompatibility, 0, len(h.cfg.OpenAICompatibility)) for _, v := range h.cfg.OpenAICompatibility { @@ -522,7 +543,7 @@ func (h *Handler) DeleteOpenAICompat(c *gin.Context) { } h.cfg.OpenAICompatibility = out h.cfg.SanitizeOpenAICompatibility() - h.persist(c) + h.persistLocked(c) return } if idxStr := c.Query("index"); idxStr != "" { @@ -531,7 +552,7 @@ func (h *Handler) DeleteOpenAICompat(c *gin.Context) { if err == nil && idx >= 0 && idx < len(h.cfg.OpenAICompatibility) { h.cfg.OpenAICompatibility = append(h.cfg.OpenAICompatibility[:idx], h.cfg.OpenAICompatibility[idx+1:]...) h.cfg.SanitizeOpenAICompatibility() - h.persist(c) + h.persistLocked(c) return } } @@ -540,7 +561,7 @@ func (h *Handler) DeleteOpenAICompat(c *gin.Context) { // vertex-api-key: []VertexCompatKey func (h *Handler) GetVertexCompatKeys(c *gin.Context) { - c.JSON(200, gin.H{"vertex-api-key": h.cfg.VertexCompatAPIKey}) + c.JSON(200, gin.H{"vertex-api-key": h.vertexCompatKeysWithAuthIndex()}) } func (h *Handler) PutVertexCompatKeys(c *gin.Context) { data, err := c.GetRawData() @@ -566,9 +587,11 @@ func (h *Handler) PutVertexCompatKeys(c *gin.Context) { return } } + h.mu.Lock() + defer h.mu.Unlock() h.cfg.VertexCompatAPIKey = append([]config.VertexCompatKey(nil), arr...) h.cfg.SanitizeVertexCompatKeys() - h.persist(c) + h.persistLocked(c) } func (h *Handler) PatchVertexCompatKey(c *gin.Context) { type vertexCompatPatch struct { @@ -589,6 +612,9 @@ func (h *Handler) PatchVertexCompatKey(c *gin.Context) { c.JSON(400, gin.H{"error": "invalid body"}) return } + + h.mu.Lock() + defer h.mu.Unlock() targetIndex := -1 if body.Index != nil && *body.Index >= 0 && *body.Index < len(h.cfg.VertexCompatAPIKey) { targetIndex = *body.Index @@ -615,7 +641,7 @@ func (h *Handler) PatchVertexCompatKey(c *gin.Context) { if trimmed == "" { h.cfg.VertexCompatAPIKey = append(h.cfg.VertexCompatAPIKey[:targetIndex], h.cfg.VertexCompatAPIKey[targetIndex+1:]...) h.cfg.SanitizeVertexCompatKeys() - h.persist(c) + h.persistLocked(c) return } entry.APIKey = trimmed @@ -628,7 +654,7 @@ func (h *Handler) PatchVertexCompatKey(c *gin.Context) { if trimmed == "" { h.cfg.VertexCompatAPIKey = append(h.cfg.VertexCompatAPIKey[:targetIndex], h.cfg.VertexCompatAPIKey[targetIndex+1:]...) h.cfg.SanitizeVertexCompatKeys() - h.persist(c) + h.persistLocked(c) return } entry.BaseURL = trimmed @@ -648,10 +674,12 @@ func (h *Handler) PatchVertexCompatKey(c *gin.Context) { normalizeVertexCompatKey(&entry) h.cfg.VertexCompatAPIKey[targetIndex] = entry h.cfg.SanitizeVertexCompatKeys() - h.persist(c) + h.persistLocked(c) } func (h *Handler) DeleteVertexCompatKey(c *gin.Context) { + h.mu.Lock() + defer h.mu.Unlock() if val := strings.TrimSpace(c.Query("api-key")); val != "" { if baseRaw, okBase := c.GetQuery("base-url"); okBase { base := strings.TrimSpace(baseRaw) @@ -664,7 +692,7 @@ func (h *Handler) DeleteVertexCompatKey(c *gin.Context) { } h.cfg.VertexCompatAPIKey = out h.cfg.SanitizeVertexCompatKeys() - h.persist(c) + h.persistLocked(c) return } @@ -686,7 +714,7 @@ func (h *Handler) DeleteVertexCompatKey(c *gin.Context) { h.cfg.VertexCompatAPIKey = append(h.cfg.VertexCompatAPIKey[:matchIndex], h.cfg.VertexCompatAPIKey[matchIndex+1:]...) } h.cfg.SanitizeVertexCompatKeys() - h.persist(c) + h.persistLocked(c) return } if idxStr := c.Query("index"); idxStr != "" { @@ -695,7 +723,7 @@ func (h *Handler) DeleteVertexCompatKey(c *gin.Context) { if errScan == nil && idx >= 0 && idx < len(h.cfg.VertexCompatAPIKey) { h.cfg.VertexCompatAPIKey = append(h.cfg.VertexCompatAPIKey[:idx], h.cfg.VertexCompatAPIKey[idx+1:]...) h.cfg.SanitizeVertexCompatKeys() - h.persist(c) + h.persistLocked(c) return } } @@ -890,7 +918,7 @@ func (h *Handler) DeleteOAuthModelAlias(c *gin.Context) { // codex-api-key: []CodexKey func (h *Handler) GetCodexKeys(c *gin.Context) { - c.JSON(200, gin.H{"codex-api-key": h.cfg.CodexKey}) + c.JSON(200, gin.H{"codex-api-key": h.codexKeysWithAuthIndex()}) } func (h *Handler) PutCodexKeys(c *gin.Context) { data, err := c.GetRawData() @@ -919,9 +947,11 @@ func (h *Handler) PutCodexKeys(c *gin.Context) { } filtered = append(filtered, entry) } + h.mu.Lock() + defer h.mu.Unlock() h.cfg.CodexKey = filtered h.cfg.SanitizeCodexKeys() - h.persist(c) + h.persistLocked(c) } func (h *Handler) PatchCodexKey(c *gin.Context) { type codexKeyPatch struct { @@ -942,6 +972,9 @@ func (h *Handler) PatchCodexKey(c *gin.Context) { c.JSON(400, gin.H{"error": "invalid body"}) return } + + h.mu.Lock() + defer h.mu.Unlock() targetIndex := -1 if body.Index != nil && *body.Index >= 0 && *body.Index < len(h.cfg.CodexKey) { targetIndex = *body.Index @@ -972,7 +1005,7 @@ func (h *Handler) PatchCodexKey(c *gin.Context) { if trimmed == "" { h.cfg.CodexKey = append(h.cfg.CodexKey[:targetIndex], h.cfg.CodexKey[targetIndex+1:]...) h.cfg.SanitizeCodexKeys() - h.persist(c) + h.persistLocked(c) return } entry.BaseURL = trimmed @@ -992,10 +1025,12 @@ func (h *Handler) PatchCodexKey(c *gin.Context) { normalizeCodexKey(&entry) h.cfg.CodexKey[targetIndex] = entry h.cfg.SanitizeCodexKeys() - h.persist(c) + h.persistLocked(c) } func (h *Handler) DeleteCodexKey(c *gin.Context) { + h.mu.Lock() + defer h.mu.Unlock() if val := strings.TrimSpace(c.Query("api-key")); val != "" { if baseRaw, okBase := c.GetQuery("base-url"); okBase { base := strings.TrimSpace(baseRaw) @@ -1008,7 +1043,7 @@ func (h *Handler) DeleteCodexKey(c *gin.Context) { } h.cfg.CodexKey = out h.cfg.SanitizeCodexKeys() - h.persist(c) + h.persistLocked(c) return } @@ -1030,7 +1065,7 @@ func (h *Handler) DeleteCodexKey(c *gin.Context) { h.cfg.CodexKey = append(h.cfg.CodexKey[:matchIndex], h.cfg.CodexKey[matchIndex+1:]...) } h.cfg.SanitizeCodexKeys() - h.persist(c) + h.persistLocked(c) return } if idxStr := c.Query("index"); idxStr != "" { @@ -1039,7 +1074,7 @@ func (h *Handler) DeleteCodexKey(c *gin.Context) { if err == nil && idx >= 0 && idx < len(h.cfg.CodexKey) { h.cfg.CodexKey = append(h.cfg.CodexKey[:idx], h.cfg.CodexKey[idx+1:]...) h.cfg.SanitizeCodexKeys() - h.persist(c) + h.persistLocked(c) return } } diff --git a/internal/api/handlers/management/handler.go b/internal/api/handlers/management/handler.go index 45786b9d3e..30cc973817 100644 --- a/internal/api/handlers/management/handler.go +++ b/internal/api/handlers/management/handler.go @@ -105,10 +105,24 @@ func NewHandlerWithoutConfigFilePath(cfg *config.Config, manager *coreauth.Manag } // SetConfig updates the in-memory config reference when the server hot-reloads. -func (h *Handler) SetConfig(cfg *config.Config) { h.cfg = cfg } +func (h *Handler) SetConfig(cfg *config.Config) { + if h == nil { + return + } + h.mu.Lock() + h.cfg = cfg + h.mu.Unlock() +} // SetAuthManager updates the auth manager reference used by management endpoints. -func (h *Handler) SetAuthManager(manager *coreauth.Manager) { h.authManager = manager } +func (h *Handler) SetAuthManager(manager *coreauth.Manager) { + if h == nil { + return + } + h.mu.Lock() + h.authManager = manager + h.mu.Unlock() +} // SetUsageStatistics allows replacing the usage statistics reference. func (h *Handler) SetUsageStatistics(stats *usage.RequestStatistics) { h.usageStats = stats } @@ -276,6 +290,12 @@ func (h *Handler) Middleware() gin.HandlerFunc { func (h *Handler) persist(c *gin.Context) bool { h.mu.Lock() defer h.mu.Unlock() + return h.persistLocked(c) +} + +// persistLocked saves the current in-memory config to disk. +// It expects the caller to hold h.mu. +func (h *Handler) persistLocked(c *gin.Context) bool { // Preserve comments when writing if err := config.SaveConfigPreserveComments(h.configFilePath, h.cfg); err != nil { c.JSON(http.StatusInternalServerError, gin.H{"error": fmt.Sprintf("failed to save config: %v", err)}) diff --git a/internal/api/server.go b/internal/api/server.go index ee24aead23..0a7fff2385 100644 --- a/internal/api/server.go +++ b/internal/api/server.go @@ -325,9 +325,16 @@ func NewServer(cfg *config.Config, authManager *auth.Manager, accessManager *sdk // setupRoutes configures the API routes for the server. // It defines the endpoints and associates them with their respective handlers. func (s *Server) setupRoutes() { - s.engine.GET("/healthz", func(c *gin.Context) { + healthzHandler := func(c *gin.Context) { + if c.Request.Method == http.MethodHead { + c.Status(http.StatusOK) + return + } + c.JSON(http.StatusOK, gin.H{"status": "ok"}) - }) + } + s.engine.GET("/healthz", healthzHandler) + s.engine.HEAD("/healthz", healthzHandler) s.engine.GET("/management.html", s.serveManagementControlPanel) openaiHandlers := openai.NewOpenAIAPIHandler(s.handlers) @@ -343,6 +350,8 @@ func (s *Server) setupRoutes() { v1.GET("/models", s.unifiedModelsHandler(openaiHandlers, claudeCodeHandlers)) v1.POST("/chat/completions", openaiHandlers.ChatCompletions) v1.POST("/completions", openaiHandlers.Completions) + v1.POST("/images/generations", openaiHandlers.ImagesGenerations) + v1.POST("/images/edits", openaiHandlers.ImagesEdits) v1.POST("/messages", claudeCodeHandlers.ClaudeMessages) v1.POST("/messages/count_tokens", claudeCodeHandlers.ClaudeCountTokens) v1.GET("/responses", openaiResponsesHandlers.ResponsesWebsocket) diff --git a/internal/api/server_test.go b/internal/api/server_test.go index dbc2cd5a83..db1ef27d17 100644 --- a/internal/api/server_test.go +++ b/internal/api/server_test.go @@ -50,23 +50,38 @@ func newTestServer(t *testing.T) *Server { func TestHealthz(t *testing.T) { server := newTestServer(t) - req := httptest.NewRequest(http.MethodGet, "/healthz", nil) - rr := httptest.NewRecorder() - server.engine.ServeHTTP(rr, req) + t.Run("GET", func(t *testing.T) { + req := httptest.NewRequest(http.MethodGet, "/healthz", nil) + rr := httptest.NewRecorder() + server.engine.ServeHTTP(rr, req) - if rr.Code != http.StatusOK { - t.Fatalf("unexpected status code: got %d want %d; body=%s", rr.Code, http.StatusOK, rr.Body.String()) - } + if rr.Code != http.StatusOK { + t.Fatalf("unexpected status code: got %d want %d; body=%s", rr.Code, http.StatusOK, rr.Body.String()) + } - var resp struct { - Status string `json:"status"` - } - if err := json.Unmarshal(rr.Body.Bytes(), &resp); err != nil { - t.Fatalf("failed to parse response JSON: %v; body=%s", err, rr.Body.String()) - } - if resp.Status != "ok" { - t.Fatalf("unexpected response status: got %q want %q", resp.Status, "ok") - } + var resp struct { + Status string `json:"status"` + } + if err := json.Unmarshal(rr.Body.Bytes(), &resp); err != nil { + t.Fatalf("failed to parse response JSON: %v; body=%s", err, rr.Body.String()) + } + if resp.Status != "ok" { + t.Fatalf("unexpected response status: got %q want %q", resp.Status, "ok") + } + }) + + t.Run("HEAD", func(t *testing.T) { + req := httptest.NewRequest(http.MethodHead, "/healthz", nil) + rr := httptest.NewRecorder() + server.engine.ServeHTTP(rr, req) + + if rr.Code != http.StatusOK { + t.Fatalf("unexpected status code: got %d want %d; body=%s", rr.Code, http.StatusOK, rr.Body.String()) + } + if rr.Body.Len() != 0 { + t.Fatalf("expected empty body for HEAD request, got %q", rr.Body.String()) + } + }) } func TestAmpProviderModelRoutes(t *testing.T) { diff --git a/internal/registry/model_definitions.go b/internal/registry/model_definitions.go index bb1206af4c..5540ca089e 100644 --- a/internal/registry/model_definitions.go +++ b/internal/registry/model_definitions.go @@ -6,6 +6,8 @@ import ( "strings" ) +const codexBuiltinImageModelID = "gpt-image-2" + // staticModelsJSON mirrors the top-level structure of models.json. type staticModelsJSON struct { Claude []*ModelInfo `json:"claude"` @@ -48,22 +50,22 @@ func GetAIStudioModels() []*ModelInfo { // GetCodexFreeModels returns model definitions for the Codex free plan tier. func GetCodexFreeModels() []*ModelInfo { - return cloneModelInfos(getModels().CodexFree) + return WithCodexBuiltins(cloneModelInfos(getModels().CodexFree)) } // GetCodexTeamModels returns model definitions for the Codex team plan tier. func GetCodexTeamModels() []*ModelInfo { - return cloneModelInfos(getModels().CodexTeam) + return WithCodexBuiltins(cloneModelInfos(getModels().CodexTeam)) } // GetCodexPlusModels returns model definitions for the Codex plus plan tier. func GetCodexPlusModels() []*ModelInfo { - return cloneModelInfos(getModels().CodexPlus) + return WithCodexBuiltins(cloneModelInfos(getModels().CodexPlus)) } // GetCodexProModels returns model definitions for the Codex pro plan tier. func GetCodexProModels() []*ModelInfo { - return cloneModelInfos(getModels().CodexPro) + return WithCodexBuiltins(cloneModelInfos(getModels().CodexPro)) } // GetKimiModels returns the standard Kimi (Moonshot AI) model definitions. @@ -272,6 +274,71 @@ func GetCodeBuddyAIModels() []*ModelInfo { } } +// WithCodexBuiltins injects hard-coded Codex-only model definitions that should +// not depend on remote models.json updates. Built-ins replace any matching IDs +// already present in the provided slice. +func WithCodexBuiltins(models []*ModelInfo) []*ModelInfo { + return upsertModelInfos(models, codexBuiltinImageModelInfo()) +} + +func codexBuiltinImageModelInfo() *ModelInfo { + return &ModelInfo{ + ID: codexBuiltinImageModelID, + Object: "model", + Created: 1704067200, // 2024-01-01 + OwnedBy: "openai", + Type: "openai", + DisplayName: "GPT Image 2", + Version: codexBuiltinImageModelID, + } +} + +func upsertModelInfos(models []*ModelInfo, extras ...*ModelInfo) []*ModelInfo { + if len(extras) == 0 { + return models + } + + extraIDs := make(map[string]struct{}, len(extras)) + extraList := make([]*ModelInfo, 0, len(extras)) + for _, extra := range extras { + if extra == nil { + continue + } + id := strings.TrimSpace(extra.ID) + if id == "" { + continue + } + key := strings.ToLower(id) + if _, exists := extraIDs[key]; exists { + continue + } + extraIDs[key] = struct{}{} + extraList = append(extraList, cloneModelInfo(extra)) + } + + if len(extraList) == 0 { + return models + } + + filtered := make([]*ModelInfo, 0, len(models)+len(extraList)) + for _, model := range models { + if model == nil { + continue + } + id := strings.TrimSpace(model.ID) + if id == "" { + continue + } + if _, exists := extraIDs[strings.ToLower(id)]; exists { + continue + } + filtered = append(filtered, model) + } + + filtered = append(filtered, extraList...) + return filtered +} + // cloneModelInfos returns a shallow copy of the slice with each element deep-cloned. func cloneModelInfos(models []*ModelInfo) []*ModelInfo { if len(models) == 0 { diff --git a/internal/registry/model_registry_safety_test.go b/internal/registry/model_registry_safety_test.go index 5f4f65d298..be5bf7908c 100644 --- a/internal/registry/model_registry_safety_test.go +++ b/internal/registry/model_registry_safety_test.go @@ -136,13 +136,13 @@ func TestGetAvailableModelsReturnsClonedSupportedParameters(t *testing.T) { } func TestLookupModelInfoReturnsCloneForStaticDefinitions(t *testing.T) { - first := LookupModelInfo("glm-4.6") + first := LookupModelInfo("claude-sonnet-4-6") if first == nil || first.Thinking == nil || len(first.Thinking.Levels) == 0 { t.Fatalf("expected static model with thinking levels, got %+v", first) } first.Thinking.Levels[0] = "mutated" - second := LookupModelInfo("glm-4.6") + second := LookupModelInfo("claude-sonnet-4-6") if second == nil || second.Thinking == nil || len(second.Thinking.Levels) == 0 || second.Thinking.Levels[0] == "mutated" { t.Fatalf("expected static lookup clone, got %+v", second) } diff --git a/internal/registry/models/models.json b/internal/registry/models/models.json index 65d8325169..24b96ca95f 100644 --- a/internal/registry/models/models.json +++ b/internal/registry/models/models.json @@ -1670,6 +1670,23 @@ "zero_allowed": true, "dynamic_allowed": true } + }, + { + "id": "kimi-k2.6", + "object": "model", + "created": 1776729600, + "owned_by": "moonshot", + "type": "kimi", + "display_name": "Kimi K2.6", + "description": "Kimi K2.6 - Latest Moonshot AI coding model with improved capabilities", + "context_length": 262144, + "max_completion_tokens": 65536, + "thinking": { + "min": 1024, + "max": 32000, + "zero_allowed": true, + "dynamic_allowed": true + } } ], "antigravity": [ diff --git a/internal/runtime/executor/claude_executor.go b/internal/runtime/executor/claude_executor.go index 51a1234860..235db1f3b2 100644 --- a/internal/runtime/executor/claude_executor.go +++ b/internal/runtime/executor/claude_executor.go @@ -11,7 +11,6 @@ import ( "fmt" "io" "net/http" - "net/textproto" "strings" "time" @@ -911,23 +910,8 @@ func applyClaudeHeaders(r *http.Request, auth *cliproxyauth.Auth, apiKey string, baseBetas += ",interleaved-thinking-2025-05-14" } - hasClaude1MHeader := false - if ginHeaders != nil { - if _, ok := ginHeaders[textproto.CanonicalMIMEHeaderKey("X-CPA-CLAUDE-1M")]; ok { - hasClaude1MHeader = true - } - } - // Also check auth attributes — GitLab Duo sets gitlab_duo_force_context_1m - // when routing through the Anthropic gateway, but the gin headers won't have - // X-CPA-CLAUDE-1M because the request is internally constructed. - if !hasClaude1MHeader && auth != nil && auth.Attributes != nil { - if auth.Attributes["gitlab_duo_force_context_1m"] == "true" { - hasClaude1MHeader = true - } - } - // Merge extra betas from request body and request flags. - if len(extraBetas) > 0 || hasClaude1MHeader { + if len(extraBetas) > 0 { existingSet := make(map[string]bool) for _, b := range strings.Split(baseBetas, ",") { betaName := strings.TrimSpace(b) @@ -942,9 +926,6 @@ func applyClaudeHeaders(r *http.Request, auth *cliproxyauth.Auth, apiKey string, existingSet[beta] = true } } - if hasClaude1MHeader && !existingSet["context-1m-2025-08-07"] { - baseBetas += ",context-1m-2025-08-07" - } } r.Header.Set("Anthropic-Beta", baseBetas) diff --git a/internal/runtime/executor/claude_executor_test.go b/internal/runtime/executor/claude_executor_test.go index f456064dc6..c1ce8fc088 100644 --- a/internal/runtime/executor/claude_executor_test.go +++ b/internal/runtime/executor/claude_executor_test.go @@ -1714,7 +1714,27 @@ func TestClaudeExecutor_ExecuteStream_AcceptEncodingOverrideCannotBypassIdentity } } -// Test case 1: String system prompt is preserved and converted to a content block +func expectedClaudeCodeStaticPrompt() string { + return strings.Join([]string{ + helps.ClaudeCodeIntro, + helps.ClaudeCodeSystem, + helps.ClaudeCodeDoingTasks, + helps.ClaudeCodeToneAndStyle, + helps.ClaudeCodeOutputEfficiency, + }, "\n\n") +} + +func expectedForwardedSystemReminder(text string) string { + return fmt.Sprintf(` +As you answer the user's questions, you can use the following context from the system: +%s + +IMPORTANT: this context may or may not be relevant to your tasks. You should not respond to this context unless it is highly relevant to your task. + +`, text) +} + +// Test case 1: String system prompt is preserved by forwarding it to the first user message func TestCheckSystemInstructionsWithMode_StringSystemPreserved(t *testing.T) { payload := []byte(`{"system":"You are a helpful assistant.","messages":[{"role":"user","content":"hi"}]}`) @@ -1733,42 +1753,52 @@ func TestCheckSystemInstructionsWithMode_StringSystemPreserved(t *testing.T) { if !strings.HasPrefix(blocks[0].Get("text").String(), "x-anthropic-billing-header:") { t.Fatalf("blocks[0] should be billing header, got %q", blocks[0].Get("text").String()) } - if blocks[1].Get("text").String() != "You are a Claude agent, built on Anthropic's Claude Agent SDK." { + if blocks[1].Get("text").String() != "You are Claude Code, Anthropic's official CLI for Claude." { t.Fatalf("blocks[1] should be agent block, got %q", blocks[1].Get("text").String()) } - if blocks[2].Get("text").String() != "You are a helpful assistant." { - t.Fatalf("blocks[2] should be user system prompt, got %q", blocks[2].Get("text").String()) + if blocks[2].Get("text").String() != expectedClaudeCodeStaticPrompt() { + t.Fatalf("blocks[2] should be static Claude Code prompt, got %q", blocks[2].Get("text").String()) + } + if blocks[2].Get("cache_control").Exists() { + t.Fatalf("blocks[2] should not have cache_control, got %s", blocks[2].Get("cache_control").Raw) } - if blocks[2].Get("cache_control.type").String() != "ephemeral" { - t.Fatalf("blocks[2] should have cache_control.type=ephemeral") + + if got := gjson.GetBytes(out, "messages.0.content").String(); got != expectedForwardedSystemReminder("You are a helpful assistant.")+"hi" { + t.Fatalf("messages[0].content should include forwarded system prompt, got %q", got) } } -// Test case 2: Strict mode drops the string system prompt +// Test case 2: Strict mode keeps only the injected Claude Code system blocks func TestCheckSystemInstructionsWithMode_StringSystemStrict(t *testing.T) { payload := []byte(`{"system":"You are a helpful assistant.","messages":[{"role":"user","content":"hi"}]}`) out := checkSystemInstructionsWithMode(payload, true) blocks := gjson.GetBytes(out, "system").Array() - if len(blocks) != 2 { - t.Fatalf("strict mode should produce 2 blocks, got %d", len(blocks)) + if len(blocks) != 3 { + t.Fatalf("strict mode should produce 3 injected blocks, got %d", len(blocks)) + } + if got := gjson.GetBytes(out, "messages.0.content").String(); got != "hi" { + t.Fatalf("strict mode should not forward system prompt into messages, got %q", got) } } -// Test case 3: Empty string system prompt does not produce a spurious block +// Test case 3: Empty string system prompt does not alter the first user message func TestCheckSystemInstructionsWithMode_EmptyStringSystemIgnored(t *testing.T) { payload := []byte(`{"system":"","messages":[{"role":"user","content":"hi"}]}`) out := checkSystemInstructionsWithMode(payload, false) blocks := gjson.GetBytes(out, "system").Array() - if len(blocks) != 2 { - t.Fatalf("empty string system should produce 2 blocks, got %d", len(blocks)) + if len(blocks) != 3 { + t.Fatalf("empty string system should still produce 3 injected blocks, got %d", len(blocks)) + } + if got := gjson.GetBytes(out, "messages.0.content").String(); got != "hi" { + t.Fatalf("empty string system should not alter messages, got %q", got) } } -// Test case 4: Array system prompt is unaffected by the string handling +// Test case 4: Array system prompt is forwarded to the first user message func TestCheckSystemInstructionsWithMode_ArraySystemStillWorks(t *testing.T) { payload := []byte(`{"system":[{"type":"text","text":"Be concise."}],"messages":[{"role":"user","content":"hi"}]}`) @@ -1778,12 +1808,15 @@ func TestCheckSystemInstructionsWithMode_ArraySystemStillWorks(t *testing.T) { if len(blocks) != 3 { t.Fatalf("expected 3 system blocks, got %d", len(blocks)) } - if blocks[2].Get("text").String() != "Be concise." { - t.Fatalf("blocks[2] should be user system prompt, got %q", blocks[2].Get("text").String()) + if blocks[2].Get("text").String() != expectedClaudeCodeStaticPrompt() { + t.Fatalf("blocks[2] should be static Claude Code prompt, got %q", blocks[2].Get("text").String()) + } + if got := gjson.GetBytes(out, "messages.0.content").String(); got != expectedForwardedSystemReminder("Be concise.")+"hi" { + t.Fatalf("messages[0].content should include forwarded array system prompt, got %q", got) } } -// Test case 5: Special characters in string system prompt survive conversion +// Test case 5: Special characters in string system prompt survive forwarding func TestCheckSystemInstructionsWithMode_StringWithSpecialChars(t *testing.T) { payload := []byte(`{"system":"Use tags & \"quotes\" in output.","messages":[{"role":"user","content":"hi"}]}`) @@ -1793,8 +1826,8 @@ func TestCheckSystemInstructionsWithMode_StringWithSpecialChars(t *testing.T) { if len(blocks) != 3 { t.Fatalf("expected 3 system blocks, got %d", len(blocks)) } - if blocks[2].Get("text").String() != `Use tags & "quotes" in output.` { - t.Fatalf("blocks[2] text mangled, got %q", blocks[2].Get("text").String()) + if got := gjson.GetBytes(out, "messages.0.content").String(); got != expectedForwardedSystemReminder(`Use tags & "quotes" in output.`)+"hi" { + t.Fatalf("forwarded system prompt text mangled, got %q", got) } } @@ -1902,8 +1935,11 @@ func TestApplyCloaking_PreservesConfiguredStrictModeAndSensitiveWordsWhenModeOmi out := applyCloaking(context.Background(), cfg, auth, payload, "claude-3-5-sonnet-20241022", "key-123") blocks := gjson.GetBytes(out, "system").Array() - if len(blocks) != 2 { - t.Fatalf("expected strict mode to keep only injected system blocks, got %d", len(blocks)) + if len(blocks) != 3 { + t.Fatalf("expected strict mode to keep the 3 injected Claude Code system blocks, got %d", len(blocks)) + } + if got := gjson.GetBytes(out, "messages.0.content.#").Int(); got != 1 { + t.Fatalf("strict mode should not prepend a forwarded system reminder block, got %d content blocks", got) } if got := gjson.GetBytes(out, "messages.0.content.0.text").String(); !strings.Contains(got, "\u200B") { t.Fatalf("expected configured sensitive word obfuscation to apply, got %q", got) diff --git a/internal/runtime/executor/codex_executor.go b/internal/runtime/executor/codex_executor.go index 41b1c32527..7d4d3edf89 100644 --- a/internal/runtime/executor/codex_executor.go +++ b/internal/runtime/executor/codex_executor.go @@ -36,6 +36,69 @@ const ( var dataTag = []byte("data:") +// Streamed Codex responses may emit response.output_item.done events while leaving +// response.completed.response.output empty. Keep the stream path aligned with the +// already-patched non-stream path by reconstructing response.output from those items. +func collectCodexOutputItemDone(eventData []byte, outputItemsByIndex map[int64][]byte, outputItemsFallback *[][]byte) { + itemResult := gjson.GetBytes(eventData, "item") + if !itemResult.Exists() || itemResult.Type != gjson.JSON { + return + } + outputIndexResult := gjson.GetBytes(eventData, "output_index") + if outputIndexResult.Exists() { + outputItemsByIndex[outputIndexResult.Int()] = []byte(itemResult.Raw) + return + } + *outputItemsFallback = append(*outputItemsFallback, []byte(itemResult.Raw)) +} + +func patchCodexCompletedOutput(eventData []byte, outputItemsByIndex map[int64][]byte, outputItemsFallback [][]byte) []byte { + outputResult := gjson.GetBytes(eventData, "response.output") + shouldPatchOutput := (!outputResult.Exists() || !outputResult.IsArray() || len(outputResult.Array()) == 0) && (len(outputItemsByIndex) > 0 || len(outputItemsFallback) > 0) + if !shouldPatchOutput { + return eventData + } + + indexes := make([]int64, 0, len(outputItemsByIndex)) + for idx := range outputItemsByIndex { + indexes = append(indexes, idx) + } + sort.Slice(indexes, func(i, j int) bool { + return indexes[i] < indexes[j] + }) + + items := make([][]byte, 0, len(outputItemsByIndex)+len(outputItemsFallback)) + for _, idx := range indexes { + items = append(items, outputItemsByIndex[idx]) + } + items = append(items, outputItemsFallback...) + + outputArray := []byte("[]") + if len(items) > 0 { + var buf bytes.Buffer + totalLen := 2 + for _, item := range items { + totalLen += len(item) + } + if len(items) > 1 { + totalLen += len(items) - 1 + } + buf.Grow(totalLen) + buf.WriteByte('[') + for i, item := range items { + if i > 0 { + buf.WriteByte(',') + } + buf.Write(item) + } + buf.WriteByte(']') + outputArray = buf.Bytes() + } + + completedDataPatched, _ := sjson.SetRawBytes(eventData, "response.output", outputArray) + return completedDataPatched +} + // CodexExecutor is a stateless executor for Codex (OpenAI Responses API entrypoint). // If api_key is unavailable on auth, it falls back to legacy via ClientAdapter. type CodexExecutor struct { @@ -414,20 +477,28 @@ func (e *CodexExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au scanner := bufio.NewScanner(httpResp.Body) scanner.Buffer(nil, 52_428_800) // 50MB var param any + outputItemsByIndex := make(map[int64][]byte) + var outputItemsFallback [][]byte for scanner.Scan() { line := scanner.Bytes() helps.AppendAPIResponseChunk(ctx, e.cfg, line) + translatedLine := bytes.Clone(line) if bytes.HasPrefix(line, dataTag) { data := bytes.TrimSpace(line[5:]) - if gjson.GetBytes(data, "type").String() == "response.completed" { + switch gjson.GetBytes(data, "type").String() { + case "response.output_item.done": + collectCodexOutputItemDone(data, outputItemsByIndex, &outputItemsFallback) + case "response.completed": if detail, ok := helps.ParseCodexUsage(data); ok { reporter.Publish(ctx, detail) } + data = patchCodexCompletedOutput(data, outputItemsByIndex, outputItemsFallback) + translatedLine = append([]byte("data: "), data...) } } - chunks := sdktranslator.TranslateStream(ctx, to, from, req.Model, originalPayload, body, bytes.Clone(line), ¶m) + chunks := sdktranslator.TranslateStream(ctx, to, from, req.Model, originalPayload, body, translatedLine, ¶m) for i := range chunks { out <- cliproxyexecutor.StreamChunk{Payload: chunks[i]} } diff --git a/internal/runtime/executor/codex_executor_stream_output_test.go b/internal/runtime/executor/codex_executor_stream_output_test.go index 91d9b0761c..a2da45e199 100644 --- a/internal/runtime/executor/codex_executor_stream_output_test.go +++ b/internal/runtime/executor/codex_executor_stream_output_test.go @@ -1,6 +1,7 @@ package executor import ( + "bytes" "context" "net/http" "net/http/httptest" @@ -44,3 +45,53 @@ func TestCodexExecutorExecute_EmptyStreamCompletionOutputUsesOutputItemDone(t *t t.Fatalf("choices.0.message.content = %q, want %q; payload=%s", gotContent, "ok", string(resp.Payload)) } } + +func TestCodexExecutorExecuteStream_EmptyStreamCompletionOutputUsesOutputItemDone(t *testing.T) { + server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) { + w.Header().Set("Content-Type", "text/event-stream") + _, _ = w.Write([]byte("data: {\"type\":\"response.output_item.done\",\"item\":{\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"output_text\",\"text\":\"ok\"}]},\"output_index\":0}\n")) + _, _ = w.Write([]byte("data: {\"type\":\"response.completed\",\"response\":{\"id\":\"resp_1\",\"object\":\"response\",\"created_at\":1775555723,\"status\":\"completed\",\"model\":\"gpt-5.4-mini-2026-03-17\",\"output\":[],\"usage\":{\"input_tokens\":8,\"output_tokens\":28,\"total_tokens\":36}}}\n\n")) + })) + defer server.Close() + + executor := NewCodexExecutor(&config.Config{}) + auth := &cliproxyauth.Auth{Attributes: map[string]string{ + "base_url": server.URL, + "api_key": "test", + }} + + result, err := executor.ExecuteStream(context.Background(), auth, cliproxyexecutor.Request{ + Model: "gpt-5.4-mini", + Payload: []byte(`{"model":"gpt-5.4-mini","input":"Say ok"}`), + }, cliproxyexecutor.Options{ + SourceFormat: sdktranslator.FromString("openai-response"), + Stream: true, + }) + if err != nil { + t.Fatalf("ExecuteStream error: %v", err) + } + + var completed []byte + for chunk := range result.Chunks { + if chunk.Err != nil { + t.Fatalf("stream chunk error: %v", chunk.Err) + } + payload := bytes.TrimSpace(chunk.Payload) + if !bytes.HasPrefix(payload, []byte("data:")) { + continue + } + data := bytes.TrimSpace(payload[5:]) + if gjson.GetBytes(data, "type").String() == "response.completed" { + completed = append([]byte(nil), data...) + } + } + + if len(completed) == 0 { + t.Fatal("missing response.completed chunk") + } + + gotContent := gjson.GetBytes(completed, "response.output.0.content.0.text").String() + if gotContent != "ok" { + t.Fatalf("response.output[0].content[0].text = %q, want %q; completed=%s", gotContent, "ok", string(completed)) + } +} diff --git a/internal/translator/codex/gemini/codex_gemini_response.go b/internal/translator/codex/gemini/codex_gemini_response.go index f6ef87710a..a2e4e20ea2 100644 --- a/internal/translator/codex/gemini/codex_gemini_response.go +++ b/internal/translator/codex/gemini/codex_gemini_response.go @@ -7,6 +7,8 @@ package gemini import ( "bytes" "context" + "crypto/sha256" + "strings" "time" translatorcommon "github.com/router-for-me/CLIProxyAPI/v6/internal/translator/common" @@ -25,6 +27,7 @@ type ConvertCodexResponseToGeminiParams struct { ResponseID string LastStorageOutput []byte HasOutputTextDelta bool + LastImageHashByID map[string][32]byte } // ConvertCodexResponseToGemini converts Codex streaming response format to Gemini format. @@ -48,6 +51,7 @@ func ConvertCodexResponseToGemini(_ context.Context, modelName string, originalR ResponseID: "", LastStorageOutput: nil, HasOutputTextDelta: false, + LastImageHashByID: make(map[string][32]byte), } } @@ -74,10 +78,63 @@ func ConvertCodexResponseToGemini(_ context.Context, modelName string, originalR template, _ = sjson.SetBytes(template, "responseId", params.ResponseID) } + if typeStr == "response.image_generation_call.partial_image" { + itemID := rootResult.Get("item_id").String() + b64 := rootResult.Get("partial_image_b64").String() + if b64 == "" { + return [][]byte{} + } + if itemID != "" { + if params.LastImageHashByID == nil { + params.LastImageHashByID = make(map[string][32]byte) + } + hash := sha256.Sum256([]byte(b64)) + if last, ok := params.LastImageHashByID[itemID]; ok && last == hash { + return [][]byte{} + } + params.LastImageHashByID[itemID] = hash + } + + outputFormat := rootResult.Get("output_format").String() + mimeType := mimeTypeFromCodexOutputFormat(outputFormat) + + part := []byte(`{"inlineData":{"data":"","mimeType":""}}`) + part, _ = sjson.SetBytes(part, "inlineData.data", b64) + part, _ = sjson.SetBytes(part, "inlineData.mimeType", mimeType) + template, _ = sjson.SetRawBytes(template, "candidates.0.content.parts.-1", part) + return [][]byte{template} + } + // Handle function call completion if typeStr == "response.output_item.done" { itemResult := rootResult.Get("item") itemType := itemResult.Get("type").String() + if itemType == "image_generation_call" { + itemID := itemResult.Get("id").String() + b64 := itemResult.Get("result").String() + if b64 == "" { + return [][]byte{} + } + if itemID != "" { + if params.LastImageHashByID == nil { + params.LastImageHashByID = make(map[string][32]byte) + } + hash := sha256.Sum256([]byte(b64)) + if last, ok := params.LastImageHashByID[itemID]; ok && last == hash { + return [][]byte{} + } + params.LastImageHashByID[itemID] = hash + } + + outputFormat := itemResult.Get("output_format").String() + mimeType := mimeTypeFromCodexOutputFormat(outputFormat) + + part := []byte(`{"inlineData":{"data":"","mimeType":""}}`) + part, _ = sjson.SetBytes(part, "inlineData.data", b64) + part, _ = sjson.SetBytes(part, "inlineData.mimeType", mimeType) + template, _ = sjson.SetRawBytes(template, "candidates.0.content.parts.-1", part) + return [][]byte{template} + } if itemType == "function_call" { // Create function call part functionCall := []byte(`{"functionCall":{"name":"","args":{}}}`) @@ -270,6 +327,20 @@ func ConvertCodexResponseToGeminiNonStream(_ context.Context, modelName string, }) } + case "image_generation_call": + flushPendingFunctionCalls() + b64 := value.Get("result").String() + if b64 == "" { + break + } + outputFormat := value.Get("output_format").String() + mimeType := mimeTypeFromCodexOutputFormat(outputFormat) + + part := []byte(`{"inlineData":{"data":"","mimeType":""}}`) + part, _ = sjson.SetBytes(part, "inlineData.data", b64) + part, _ = sjson.SetBytes(part, "inlineData.mimeType", mimeType) + template, _ = sjson.SetRawBytes(template, "candidates.0.content.parts.-1", part) + case "function_call": // Collect function call for potential merging with consecutive ones hasToolCall = true @@ -342,3 +413,24 @@ func buildReverseMapFromGeminiOriginal(original []byte) map[string]string { func GeminiTokenCount(ctx context.Context, count int64) []byte { return translatorcommon.GeminiTokenCountJSON(count) } + +func mimeTypeFromCodexOutputFormat(outputFormat string) string { + if outputFormat == "" { + return "image/png" + } + if strings.Contains(outputFormat, "/") { + return outputFormat + } + switch strings.ToLower(outputFormat) { + case "png": + return "image/png" + case "jpg", "jpeg": + return "image/jpeg" + case "webp": + return "image/webp" + case "gif": + return "image/gif" + default: + return "image/png" + } +} diff --git a/internal/translator/codex/gemini/codex_gemini_response_test.go b/internal/translator/codex/gemini/codex_gemini_response_test.go index b8f227beb5..547ee84715 100644 --- a/internal/translator/codex/gemini/codex_gemini_response_test.go +++ b/internal/translator/codex/gemini/codex_gemini_response_test.go @@ -33,3 +33,79 @@ func TestConvertCodexResponseToGemini_StreamEmptyOutputUsesOutputItemDoneMessage t.Fatalf("expected fallback content from response.output_item.done message; outputs=%q", outputs) } } + +func TestConvertCodexResponseToGemini_StreamPartialImageEmitsInlineData(t *testing.T) { + ctx := context.Background() + originalRequest := []byte(`{"tools":[]}`) + var param any + + chunk := []byte(`data: {"type":"response.image_generation_call.partial_image","item_id":"ig_123","output_format":"png","partial_image_b64":"aGVsbG8=","partial_image_index":0}`) + out := ConvertCodexResponseToGemini(ctx, "gemini-2.5-pro", originalRequest, nil, chunk, ¶m) + if len(out) != 1 { + t.Fatalf("expected 1 chunk, got %d", len(out)) + } + + got := gjson.GetBytes(out[0], "candidates.0.content.parts.0.inlineData.data").String() + if got != "aGVsbG8=" { + t.Fatalf("expected inlineData.data %q, got %q; chunk=%s", "aGVsbG8=", got, string(out[0])) + } + + gotMime := gjson.GetBytes(out[0], "candidates.0.content.parts.0.inlineData.mimeType").String() + if gotMime != "image/png" { + t.Fatalf("expected inlineData.mimeType %q, got %q; chunk=%s", "image/png", gotMime, string(out[0])) + } + + out = ConvertCodexResponseToGemini(ctx, "gemini-2.5-pro", originalRequest, nil, chunk, ¶m) + if len(out) != 0 { + t.Fatalf("expected duplicate image chunk to be suppressed, got %d", len(out)) + } +} + +func TestConvertCodexResponseToGemini_StreamImageGenerationCallDoneEmitsInlineData(t *testing.T) { + ctx := context.Background() + originalRequest := []byte(`{"tools":[]}`) + var param any + + out := ConvertCodexResponseToGemini(ctx, "gemini-2.5-pro", originalRequest, nil, []byte(`data: {"type":"response.image_generation_call.partial_image","item_id":"ig_123","output_format":"png","partial_image_b64":"aGVsbG8=","partial_image_index":0}`), ¶m) + if len(out) != 1 { + t.Fatalf("expected 1 chunk, got %d", len(out)) + } + + out = ConvertCodexResponseToGemini(ctx, "gemini-2.5-pro", originalRequest, nil, []byte(`data: {"type":"response.output_item.done","item":{"id":"ig_123","type":"image_generation_call","output_format":"png","result":"aGVsbG8="}}`), ¶m) + if len(out) != 0 { + t.Fatalf("expected output_item.done to be suppressed when identical to last partial image, got %d", len(out)) + } + + out = ConvertCodexResponseToGemini(ctx, "gemini-2.5-pro", originalRequest, nil, []byte(`data: {"type":"response.output_item.done","item":{"id":"ig_123","type":"image_generation_call","output_format":"jpeg","result":"Ymll"}}`), ¶m) + if len(out) != 1 { + t.Fatalf("expected 1 chunk, got %d", len(out)) + } + + got := gjson.GetBytes(out[0], "candidates.0.content.parts.0.inlineData.data").String() + if got != "Ymll" { + t.Fatalf("expected inlineData.data %q, got %q; chunk=%s", "Ymll", got, string(out[0])) + } + + gotMime := gjson.GetBytes(out[0], "candidates.0.content.parts.0.inlineData.mimeType").String() + if gotMime != "image/jpeg" { + t.Fatalf("expected inlineData.mimeType %q, got %q; chunk=%s", "image/jpeg", gotMime, string(out[0])) + } +} + +func TestConvertCodexResponseToGemini_NonStreamImageGenerationCallAddsInlineDataPart(t *testing.T) { + ctx := context.Background() + originalRequest := []byte(`{"tools":[]}`) + + raw := []byte(`{"type":"response.completed","response":{"id":"resp_123","created_at":1700000000,"usage":{"input_tokens":1,"output_tokens":1},"output":[{"type":"message","content":[{"type":"output_text","text":"ok"}]},{"type":"image_generation_call","output_format":"png","result":"aGVsbG8="}]}}`) + out := ConvertCodexResponseToGeminiNonStream(ctx, "gemini-2.5-pro", originalRequest, nil, raw, nil) + + got := gjson.GetBytes(out, "candidates.0.content.parts.1.inlineData.data").String() + if got != "aGVsbG8=" { + t.Fatalf("expected inlineData.data %q, got %q; chunk=%s", "aGVsbG8=", got, string(out)) + } + + gotMime := gjson.GetBytes(out, "candidates.0.content.parts.1.inlineData.mimeType").String() + if gotMime != "image/png" { + t.Fatalf("expected inlineData.mimeType %q, got %q; chunk=%s", "image/png", gotMime, string(out)) + } +} diff --git a/internal/translator/codex/openai/chat-completions/codex_openai_response.go b/internal/translator/codex/openai/chat-completions/codex_openai_response.go index afae35d48d..75b5b848b3 100644 --- a/internal/translator/codex/openai/chat-completions/codex_openai_response.go +++ b/internal/translator/codex/openai/chat-completions/codex_openai_response.go @@ -8,6 +8,8 @@ package chat_completions import ( "bytes" "context" + "crypto/sha256" + "strings" "time" "github.com/tidwall/gjson" @@ -26,6 +28,7 @@ type ConvertCliToOpenAIParams struct { FunctionCallIndex int HasReceivedArgumentsDelta bool HasToolCallAnnounced bool + LastImageHashByItemID map[string][32]byte } // ConvertCodexResponseToOpenAI translates a single chunk of a streaming response from the @@ -51,6 +54,7 @@ func ConvertCodexResponseToOpenAI(_ context.Context, modelName string, originalR FunctionCallIndex: -1, HasReceivedArgumentsDelta: false, HasToolCallAnnounced: false, + LastImageHashByItemID: make(map[string][32]byte), } } @@ -70,6 +74,9 @@ func ConvertCodexResponseToOpenAI(_ context.Context, modelName string, originalR (*param).(*ConvertCliToOpenAIParams).ResponseID = rootResult.Get("response.id").String() (*param).(*ConvertCliToOpenAIParams).CreatedAt = rootResult.Get("response.created_at").Int() (*param).(*ConvertCliToOpenAIParams).Model = rootResult.Get("response.model").String() + if (*param).(*ConvertCliToOpenAIParams).LastImageHashByItemID == nil { + (*param).(*ConvertCliToOpenAIParams).LastImageHashByItemID = make(map[string][32]byte) + } return [][]byte{} } @@ -120,6 +127,39 @@ func ConvertCodexResponseToOpenAI(_ context.Context, modelName string, originalR template, _ = sjson.SetBytes(template, "choices.0.delta.role", "assistant") template, _ = sjson.SetBytes(template, "choices.0.delta.content", deltaResult.String()) } + } else if dataType == "response.image_generation_call.partial_image" { + itemID := rootResult.Get("item_id").String() + b64 := rootResult.Get("partial_image_b64").String() + if b64 == "" { + return [][]byte{} + } + if itemID != "" { + p := (*param).(*ConvertCliToOpenAIParams) + if p.LastImageHashByItemID == nil { + p.LastImageHashByItemID = make(map[string][32]byte) + } + hash := sha256.Sum256([]byte(b64)) + if last, ok := p.LastImageHashByItemID[itemID]; ok && last == hash { + return [][]byte{} + } + p.LastImageHashByItemID[itemID] = hash + } + + outputFormat := rootResult.Get("output_format").String() + mimeType := mimeTypeFromCodexOutputFormat(outputFormat) + imageURL := "data:" + mimeType + ";base64," + b64 + + imagesResult := gjson.GetBytes(template, "choices.0.delta.images") + if !imagesResult.Exists() || !imagesResult.IsArray() { + template, _ = sjson.SetRawBytes(template, "choices.0.delta.images", []byte(`[]`)) + } + imageIndex := len(gjson.GetBytes(template, "choices.0.delta.images").Array()) + imagePayload := []byte(`{"type":"image_url","image_url":{"url":""}}`) + imagePayload, _ = sjson.SetBytes(imagePayload, "index", imageIndex) + imagePayload, _ = sjson.SetBytes(imagePayload, "image_url.url", imageURL) + + template, _ = sjson.SetBytes(template, "choices.0.delta.role", "assistant") + template, _ = sjson.SetRawBytes(template, "choices.0.delta.images.-1", imagePayload) } else if dataType == "response.completed" { finishReason := "stop" if (*param).(*ConvertCliToOpenAIParams).FunctionCallIndex != -1 { @@ -183,7 +223,46 @@ func ConvertCodexResponseToOpenAI(_ context.Context, modelName string, originalR } else if dataType == "response.output_item.done" { itemResult := rootResult.Get("item") - if !itemResult.Exists() || itemResult.Get("type").String() != "function_call" { + if !itemResult.Exists() { + return [][]byte{} + } + itemType := itemResult.Get("type").String() + if itemType == "image_generation_call" { + itemID := itemResult.Get("id").String() + b64 := itemResult.Get("result").String() + if b64 == "" { + return [][]byte{} + } + if itemID != "" { + p := (*param).(*ConvertCliToOpenAIParams) + if p.LastImageHashByItemID == nil { + p.LastImageHashByItemID = make(map[string][32]byte) + } + hash := sha256.Sum256([]byte(b64)) + if last, ok := p.LastImageHashByItemID[itemID]; ok && last == hash { + return [][]byte{} + } + p.LastImageHashByItemID[itemID] = hash + } + + outputFormat := itemResult.Get("output_format").String() + mimeType := mimeTypeFromCodexOutputFormat(outputFormat) + imageURL := "data:" + mimeType + ";base64," + b64 + + imagesResult := gjson.GetBytes(template, "choices.0.delta.images") + if !imagesResult.Exists() || !imagesResult.IsArray() { + template, _ = sjson.SetRawBytes(template, "choices.0.delta.images", []byte(`[]`)) + } + imageIndex := len(gjson.GetBytes(template, "choices.0.delta.images").Array()) + imagePayload := []byte(`{"type":"image_url","image_url":{"url":""}}`) + imagePayload, _ = sjson.SetBytes(imagePayload, "index", imageIndex) + imagePayload, _ = sjson.SetBytes(imagePayload, "image_url.url", imageURL) + + template, _ = sjson.SetBytes(template, "choices.0.delta.role", "assistant") + template, _ = sjson.SetRawBytes(template, "choices.0.delta.images.-1", imagePayload) + return [][]byte{template} + } + if itemType != "function_call" { return [][]byte{} } @@ -285,6 +364,7 @@ func ConvertCodexResponseToOpenAINonStream(_ context.Context, _ string, original // Process the output array for content and function calls var toolCalls [][]byte + var images [][]byte outputResult := responseResult.Get("output") if outputResult.IsArray() { outputArray := outputResult.Array() @@ -339,6 +419,19 @@ func ConvertCodexResponseToOpenAINonStream(_ context.Context, _ string, original } toolCalls = append(toolCalls, functionCallTemplate) + case "image_generation_call": + b64 := outputItem.Get("result").String() + if b64 == "" { + break + } + outputFormat := outputItem.Get("output_format").String() + mimeType := mimeTypeFromCodexOutputFormat(outputFormat) + imageURL := "data:" + mimeType + ";base64," + b64 + + imagePayload := []byte(`{"type":"image_url","image_url":{"url":""}}`) + imagePayload, _ = sjson.SetBytes(imagePayload, "index", len(images)) + imagePayload, _ = sjson.SetBytes(imagePayload, "image_url.url", imageURL) + images = append(images, imagePayload) } } @@ -361,6 +454,15 @@ func ConvertCodexResponseToOpenAINonStream(_ context.Context, _ string, original } template, _ = sjson.SetBytes(template, "choices.0.message.role", "assistant") } + + // Add images if any + if len(images) > 0 { + template, _ = sjson.SetRawBytes(template, "choices.0.message.images", []byte(`[]`)) + for _, image := range images { + template, _ = sjson.SetRawBytes(template, "choices.0.message.images.-1", image) + } + template, _ = sjson.SetBytes(template, "choices.0.message.role", "assistant") + } } // Extract and set the finish reason based on status @@ -409,3 +511,24 @@ func buildReverseMapFromOriginalOpenAI(original []byte) map[string]string { } return rev } + +func mimeTypeFromCodexOutputFormat(outputFormat string) string { + if outputFormat == "" { + return "image/png" + } + if strings.Contains(outputFormat, "/") { + return outputFormat + } + switch strings.ToLower(outputFormat) { + case "png": + return "image/png" + case "jpg", "jpeg": + return "image/jpeg" + case "webp": + return "image/webp" + case "gif": + return "image/gif" + default: + return "image/png" + } +} diff --git a/internal/translator/codex/openai/chat-completions/codex_openai_response_test.go b/internal/translator/codex/openai/chat-completions/codex_openai_response_test.go index 534884c229..a6bb486fdf 100644 --- a/internal/translator/codex/openai/chat-completions/codex_openai_response_test.go +++ b/internal/translator/codex/openai/chat-completions/codex_openai_response_test.go @@ -90,3 +90,62 @@ func TestConvertCodexResponseToOpenAI_ToolCallArgumentsDeltaOmitsNullContentFiel t.Fatalf("expected tool call arguments delta to exist, got %s", string(out[0])) } } + +func TestConvertCodexResponseToOpenAI_StreamPartialImageEmitsDeltaImages(t *testing.T) { + ctx := context.Background() + var param any + + chunk := []byte(`data: {"type":"response.image_generation_call.partial_image","item_id":"ig_123","output_format":"png","partial_image_b64":"aGVsbG8=","partial_image_index":0}`) + + out := ConvertCodexResponseToOpenAI(ctx, "gpt-5.4", nil, nil, chunk, ¶m) + if len(out) != 1 { + t.Fatalf("expected 1 chunk, got %d", len(out)) + } + + gotURL := gjson.GetBytes(out[0], "choices.0.delta.images.0.image_url.url").String() + if gotURL != "data:image/png;base64,aGVsbG8=" { + t.Fatalf("expected image url %q, got %q; chunk=%s", "data:image/png;base64,aGVsbG8=", gotURL, string(out[0])) + } + + out = ConvertCodexResponseToOpenAI(ctx, "gpt-5.4", nil, nil, chunk, ¶m) + if len(out) != 0 { + t.Fatalf("expected duplicate image chunk to be suppressed, got %d", len(out)) + } +} + +func TestConvertCodexResponseToOpenAI_StreamImageGenerationCallDoneEmitsDeltaImages(t *testing.T) { + ctx := context.Background() + var param any + + out := ConvertCodexResponseToOpenAI(ctx, "gpt-5.4", nil, nil, []byte(`data: {"type":"response.image_generation_call.partial_image","item_id":"ig_123","output_format":"png","partial_image_b64":"aGVsbG8=","partial_image_index":0}`), ¶m) + if len(out) != 1 { + t.Fatalf("expected 1 chunk, got %d", len(out)) + } + + out = ConvertCodexResponseToOpenAI(ctx, "gpt-5.4", nil, nil, []byte(`data: {"type":"response.output_item.done","item":{"id":"ig_123","type":"image_generation_call","output_format":"png","result":"aGVsbG8="}}`), ¶m) + if len(out) != 0 { + t.Fatalf("expected output_item.done to be suppressed when identical to last partial image, got %d", len(out)) + } + + out = ConvertCodexResponseToOpenAI(ctx, "gpt-5.4", nil, nil, []byte(`data: {"type":"response.output_item.done","item":{"id":"ig_123","type":"image_generation_call","output_format":"jpeg","result":"Ymll"}}`), ¶m) + if len(out) != 1 { + t.Fatalf("expected 1 chunk, got %d", len(out)) + } + + gotURL := gjson.GetBytes(out[0], "choices.0.delta.images.0.image_url.url").String() + if gotURL != "data:image/jpeg;base64,Ymll" { + t.Fatalf("expected image url %q, got %q; chunk=%s", "data:image/jpeg;base64,Ymll", gotURL, string(out[0])) + } +} + +func TestConvertCodexResponseToOpenAI_NonStreamImageGenerationCallAddsMessageImages(t *testing.T) { + ctx := context.Background() + + raw := []byte(`{"type":"response.completed","response":{"id":"resp_123","created_at":1700000000,"model":"gpt-5.4","status":"completed","usage":{"input_tokens":1,"output_tokens":1,"total_tokens":2},"output":[{"type":"message","content":[{"type":"output_text","text":"ok"}]},{"type":"image_generation_call","output_format":"png","result":"aGVsbG8="}]}}`) + out := ConvertCodexResponseToOpenAINonStream(ctx, "gpt-5.4", nil, nil, raw, nil) + + gotURL := gjson.GetBytes(out, "choices.0.message.images.0.image_url.url").String() + if gotURL != "data:image/png;base64,aGVsbG8=" { + t.Fatalf("expected image url %q, got %q; chunk=%s", "data:image/png;base64,aGVsbG8=", gotURL, string(out)) + } +} diff --git a/internal/util/header_helpers.go b/internal/util/header_helpers.go index c53c291f10..0b8d72bcb4 100644 --- a/internal/util/header_helpers.go +++ b/internal/util/header_helpers.go @@ -47,6 +47,14 @@ func applyCustomHeaders(r *http.Request, headers map[string]string) { if k == "" || v == "" { continue } + // net/http reads Host from req.Host (not req.Header) when writing + // a real request, so we must mirror it there. Some callers pass + // synthetic requests (e.g. &http.Request{Header: ...}) and only + // consume r.Header afterwards, so keep the value in the header + // map too. + if http.CanonicalHeaderKey(k) == "Host" { + r.Host = v + } r.Header.Set(k, v) } } diff --git a/sdk/api/handlers/handlers.go b/sdk/api/handlers/handlers.go index 5c416219f2..dedecab4eb 100644 --- a/sdk/api/handlers/handlers.go +++ b/sdk/api/handlers/handlers.go @@ -796,6 +796,13 @@ func (h *BaseAPIHandler) getRequestDetails(modelName string) (providers []string parsed := thinking.ParseSuffix(resolvedModelName) baseModel := strings.TrimSpace(parsed.ModelName) + if strings.EqualFold(baseModel, "gpt-image-2") { + return nil, "", &interfaces.ErrorMessage{ + StatusCode: http.StatusServiceUnavailable, + Error: fmt.Errorf("model %s is only supported on /v1/images/generations and /v1/images/edits", baseModel), + } + } + providers = util.GetProviderName(baseModel) // Fallback: if baseModel has no provider but differs from resolvedModelName, // try using the full model name. This handles edge cases where custom models diff --git a/sdk/api/handlers/handlers_request_details_test.go b/sdk/api/handlers/handlers_request_details_test.go index b0f6b13262..c98580f224 100644 --- a/sdk/api/handlers/handlers_request_details_test.go +++ b/sdk/api/handlers/handlers_request_details_test.go @@ -1,7 +1,9 @@ package handlers import ( + "net/http" "reflect" + "strings" "testing" "time" @@ -116,3 +118,22 @@ func TestGetRequestDetails_PreservesSuffix(t *testing.T) { }) } } + +func TestGetRequestDetails_ImageModelReturns503(t *testing.T) { + handler := NewBaseAPIHandlers(&sdkconfig.SDKConfig{}, coreauth.NewManager(nil, nil, nil)) + + _, _, errMsg := handler.getRequestDetails("gpt-image-2") + if errMsg == nil { + t.Fatalf("expected error for gpt-image-2, got nil") + } + if errMsg.StatusCode != http.StatusServiceUnavailable { + t.Fatalf("unexpected status code: got %d want %d", errMsg.StatusCode, http.StatusServiceUnavailable) + } + if errMsg.Error == nil { + t.Fatalf("expected error message, got nil") + } + msg := errMsg.Error.Error() + if !strings.Contains(msg, "/v1/images/generations") || !strings.Contains(msg, "/v1/images/edits") { + t.Fatalf("unexpected error message: %q", msg) + } +} diff --git a/sdk/api/handlers/openai/openai_images_handlers.go b/sdk/api/handlers/openai/openai_images_handlers.go new file mode 100644 index 0000000000..93d45460d0 --- /dev/null +++ b/sdk/api/handlers/openai/openai_images_handlers.go @@ -0,0 +1,896 @@ +package openai + +import ( + "bytes" + "context" + "encoding/base64" + "encoding/json" + "fmt" + "io" + "mime/multipart" + "net/http" + "strconv" + "strings" + "time" + + "github.com/gin-gonic/gin" + "github.com/router-for-me/CLIProxyAPI/v6/internal/interfaces" + "github.com/router-for-me/CLIProxyAPI/v6/sdk/api/handlers" + log "github.com/sirupsen/logrus" + "github.com/tidwall/gjson" + "github.com/tidwall/sjson" +) + +const ( + defaultImagesMainModel = "gpt-5.4-mini" + defaultImagesToolModel = "gpt-image-2" +) + +type imageCallResult struct { + Result string + RevisedPrompt string + OutputFormat string + Size string + Background string + Quality string +} + +type sseFrameAccumulator struct { + pending []byte +} + +func (a *sseFrameAccumulator) AddChunk(chunk []byte) [][]byte { + if len(chunk) == 0 { + return nil + } + + if responsesSSENeedsLineBreak(a.pending, chunk) { + a.pending = append(a.pending, '\n') + } + a.pending = append(a.pending, chunk...) + + var frames [][]byte + for { + frameLen := responsesSSEFrameLen(a.pending) + if frameLen == 0 { + break + } + frames = append(frames, a.pending[:frameLen]) + copy(a.pending, a.pending[frameLen:]) + a.pending = a.pending[:len(a.pending)-frameLen] + } + + if len(bytes.TrimSpace(a.pending)) == 0 { + a.pending = a.pending[:0] + return frames + } + if len(a.pending) == 0 || !responsesSSECanEmitWithoutDelimiter(a.pending) { + return frames + } + frames = append(frames, a.pending) + a.pending = a.pending[:0] + return frames +} + +func (a *sseFrameAccumulator) Flush() [][]byte { + if len(a.pending) == 0 { + return nil + } + + var frames [][]byte + for { + frameLen := responsesSSEFrameLen(a.pending) + if frameLen == 0 { + break + } + frames = append(frames, a.pending[:frameLen]) + copy(a.pending, a.pending[frameLen:]) + a.pending = a.pending[:len(a.pending)-frameLen] + } + + if len(bytes.TrimSpace(a.pending)) == 0 { + a.pending = nil + return frames + } + if responsesSSECanEmitWithoutDelimiter(a.pending) { + frames = append(frames, a.pending) + } + a.pending = nil + return frames +} + +func mimeTypeFromOutputFormat(outputFormat string) string { + if outputFormat == "" { + return "image/png" + } + if strings.Contains(outputFormat, "/") { + return outputFormat + } + switch strings.ToLower(strings.TrimSpace(outputFormat)) { + case "png": + return "image/png" + case "jpg", "jpeg": + return "image/jpeg" + case "webp": + return "image/webp" + default: + return "image/png" + } +} + +func multipartFileToDataURL(fileHeader *multipart.FileHeader) (string, error) { + if fileHeader == nil { + return "", fmt.Errorf("upload file is nil") + } + f, err := fileHeader.Open() + if err != nil { + return "", fmt.Errorf("open upload file failed: %w", err) + } + defer func() { + if errClose := f.Close(); errClose != nil { + log.Errorf("openai images: close upload file error: %v", errClose) + } + }() + + data, err := io.ReadAll(f) + if err != nil { + return "", fmt.Errorf("read upload file failed: %w", err) + } + + mediaType := strings.TrimSpace(fileHeader.Header.Get("Content-Type")) + if mediaType == "" { + mediaType = http.DetectContentType(data) + } + + b64 := base64.StdEncoding.EncodeToString(data) + return "data:" + mediaType + ";base64," + b64, nil +} + +func parseIntField(raw string, fallback int64) int64 { + raw = strings.TrimSpace(raw) + if raw == "" { + return fallback + } + v, err := strconv.ParseInt(raw, 10, 64) + if err != nil { + return fallback + } + return v +} + +func parseBoolField(raw string, fallback bool) bool { + raw = strings.TrimSpace(strings.ToLower(raw)) + if raw == "" { + return fallback + } + switch raw { + case "1", "true", "yes", "on": + return true + case "0", "false", "no", "off": + return false + default: + return fallback + } +} + +func (h *OpenAIAPIHandler) ImagesGenerations(c *gin.Context) { + rawJSON, err := c.GetRawData() + if err != nil { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: fmt.Sprintf("Invalid request: %v", err), + Type: "invalid_request_error", + }, + }) + return + } + if !json.Valid(rawJSON) { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: "Invalid request: body must be valid JSON", + Type: "invalid_request_error", + }, + }) + return + } + + prompt := strings.TrimSpace(gjson.GetBytes(rawJSON, "prompt").String()) + if prompt == "" { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: "Invalid request: prompt is required", + Type: "invalid_request_error", + }, + }) + return + } + + imageModel := strings.TrimSpace(gjson.GetBytes(rawJSON, "model").String()) + if imageModel == "" { + imageModel = defaultImagesToolModel + } + responseFormat := strings.TrimSpace(gjson.GetBytes(rawJSON, "response_format").String()) + if responseFormat == "" { + responseFormat = "b64_json" + } + stream := gjson.GetBytes(rawJSON, "stream").Bool() + + tool := []byte(`{"type":"image_generation","action":"generate"}`) + tool, _ = sjson.SetBytes(tool, "model", imageModel) + + if v := strings.TrimSpace(gjson.GetBytes(rawJSON, "size").String()); v != "" { + tool, _ = sjson.SetBytes(tool, "size", v) + } + if v := strings.TrimSpace(gjson.GetBytes(rawJSON, "quality").String()); v != "" { + tool, _ = sjson.SetBytes(tool, "quality", v) + } + if v := strings.TrimSpace(gjson.GetBytes(rawJSON, "background").String()); v != "" { + tool, _ = sjson.SetBytes(tool, "background", v) + } + if v := strings.TrimSpace(gjson.GetBytes(rawJSON, "output_format").String()); v != "" { + tool, _ = sjson.SetBytes(tool, "output_format", v) + } + if v := gjson.GetBytes(rawJSON, "output_compression"); v.Exists() { + if v.Type == gjson.Number { + tool, _ = sjson.SetBytes(tool, "output_compression", v.Int()) + } + } + if v := gjson.GetBytes(rawJSON, "partial_images"); v.Exists() { + if v.Type == gjson.Number { + tool, _ = sjson.SetBytes(tool, "partial_images", v.Int()) + } + } + if v := strings.TrimSpace(gjson.GetBytes(rawJSON, "moderation").String()); v != "" { + tool, _ = sjson.SetBytes(tool, "moderation", v) + } + + responsesReq := buildImagesResponsesRequest(prompt, nil, tool) + if stream { + h.streamImagesFromResponses(c, responsesReq, responseFormat, "image_generation") + return + } + h.collectImagesFromResponses(c, responsesReq, responseFormat) +} + +func (h *OpenAIAPIHandler) ImagesEdits(c *gin.Context) { + contentType := strings.ToLower(strings.TrimSpace(c.GetHeader("Content-Type"))) + if strings.HasPrefix(contentType, "application/json") { + h.imagesEditsFromJSON(c) + return + } + if strings.HasPrefix(contentType, "multipart/form-data") || contentType == "" { + h.imagesEditsFromMultipart(c) + return + } + + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: fmt.Sprintf("Invalid request: unsupported Content-Type %q", contentType), + Type: "invalid_request_error", + }, + }) +} + +func (h *OpenAIAPIHandler) imagesEditsFromMultipart(c *gin.Context) { + form, err := c.MultipartForm() + if err != nil { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: fmt.Sprintf("Invalid request: %v", err), + Type: "invalid_request_error", + }, + }) + return + } + + prompt := strings.TrimSpace(c.PostForm("prompt")) + if prompt == "" { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: "Invalid request: prompt is required", + Type: "invalid_request_error", + }, + }) + return + } + + var imageFiles []*multipart.FileHeader + if files := form.File["image[]"]; len(files) > 0 { + imageFiles = files + } else if files := form.File["image"]; len(files) > 0 { + imageFiles = files + } + if len(imageFiles) == 0 { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: "Invalid request: image is required", + Type: "invalid_request_error", + }, + }) + return + } + + images := make([]string, 0, len(imageFiles)) + for _, fh := range imageFiles { + dataURL, err := multipartFileToDataURL(fh) + if err != nil { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: fmt.Sprintf("Invalid request: %v", err), + Type: "invalid_request_error", + }, + }) + return + } + images = append(images, dataURL) + } + + var maskDataURL *string + if maskFiles := form.File["mask"]; len(maskFiles) > 0 && maskFiles[0] != nil { + dataURL, err := multipartFileToDataURL(maskFiles[0]) + if err != nil { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: fmt.Sprintf("Invalid request: %v", err), + Type: "invalid_request_error", + }, + }) + return + } + maskDataURL = &dataURL + } + + imageModel := strings.TrimSpace(c.PostForm("model")) + if imageModel == "" { + imageModel = defaultImagesToolModel + } + responseFormat := strings.TrimSpace(c.PostForm("response_format")) + if responseFormat == "" { + responseFormat = "b64_json" + } + stream := parseBoolField(c.PostForm("stream"), false) + + tool := []byte(`{"type":"image_generation","action":"edit"}`) + tool, _ = sjson.SetBytes(tool, "model", imageModel) + + if v := strings.TrimSpace(c.PostForm("size")); v != "" { + tool, _ = sjson.SetBytes(tool, "size", v) + } + if v := strings.TrimSpace(c.PostForm("quality")); v != "" { + tool, _ = sjson.SetBytes(tool, "quality", v) + } + if v := strings.TrimSpace(c.PostForm("background")); v != "" { + tool, _ = sjson.SetBytes(tool, "background", v) + } + if v := strings.TrimSpace(c.PostForm("output_format")); v != "" { + tool, _ = sjson.SetBytes(tool, "output_format", v) + } + if v := strings.TrimSpace(c.PostForm("input_fidelity")); v != "" { + tool, _ = sjson.SetBytes(tool, "input_fidelity", v) + } + if v := strings.TrimSpace(c.PostForm("moderation")); v != "" { + tool, _ = sjson.SetBytes(tool, "moderation", v) + } + + if v := strings.TrimSpace(c.PostForm("output_compression")); v != "" { + tool, _ = sjson.SetBytes(tool, "output_compression", parseIntField(v, 0)) + } + if v := strings.TrimSpace(c.PostForm("partial_images")); v != "" { + tool, _ = sjson.SetBytes(tool, "partial_images", parseIntField(v, 0)) + } + + if maskDataURL != nil && strings.TrimSpace(*maskDataURL) != "" { + tool, _ = sjson.SetBytes(tool, "input_image_mask.image_url", strings.TrimSpace(*maskDataURL)) + } + + responsesReq := buildImagesResponsesRequest(prompt, images, tool) + if stream { + h.streamImagesFromResponses(c, responsesReq, responseFormat, "image_edit") + return + } + h.collectImagesFromResponses(c, responsesReq, responseFormat) +} + +func (h *OpenAIAPIHandler) imagesEditsFromJSON(c *gin.Context) { + rawJSON, err := c.GetRawData() + if err != nil { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: fmt.Sprintf("Invalid request: %v", err), + Type: "invalid_request_error", + }, + }) + return + } + if !json.Valid(rawJSON) { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: "Invalid request: body must be valid JSON", + Type: "invalid_request_error", + }, + }) + return + } + + prompt := strings.TrimSpace(gjson.GetBytes(rawJSON, "prompt").String()) + if prompt == "" { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: "Invalid request: prompt is required", + Type: "invalid_request_error", + }, + }) + return + } + + var images []string + imagesResult := gjson.GetBytes(rawJSON, "images") + if imagesResult.IsArray() { + for _, img := range imagesResult.Array() { + url := strings.TrimSpace(img.Get("image_url").String()) + if url == "" { + continue + } + images = append(images, url) + } + } + if len(images) == 0 { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: "Invalid request: images[].image_url is required (file_id is not supported)", + Type: "invalid_request_error", + }, + }) + return + } + + var maskDataURL *string + if mask := gjson.GetBytes(rawJSON, "mask.image_url"); mask.Exists() { + url := strings.TrimSpace(mask.String()) + if url != "" { + maskDataURL = &url + } + } else if mask := gjson.GetBytes(rawJSON, "mask.file_id"); mask.Exists() { + c.JSON(http.StatusBadRequest, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: "Invalid request: mask.file_id is not supported (use mask.image_url instead)", + Type: "invalid_request_error", + }, + }) + return + } + + imageModel := strings.TrimSpace(gjson.GetBytes(rawJSON, "model").String()) + if imageModel == "" { + imageModel = defaultImagesToolModel + } + responseFormat := strings.TrimSpace(gjson.GetBytes(rawJSON, "response_format").String()) + if responseFormat == "" { + responseFormat = "b64_json" + } + stream := gjson.GetBytes(rawJSON, "stream").Bool() + + tool := []byte(`{"type":"image_generation","action":"edit"}`) + tool, _ = sjson.SetBytes(tool, "model", imageModel) + + for _, field := range []string{"size", "quality", "background", "output_format", "input_fidelity", "moderation"} { + if v := strings.TrimSpace(gjson.GetBytes(rawJSON, field).String()); v != "" { + tool, _ = sjson.SetBytes(tool, field, v) + } + } + + for _, field := range []string{"output_compression", "partial_images"} { + if v := gjson.GetBytes(rawJSON, field); v.Exists() && v.Type == gjson.Number { + tool, _ = sjson.SetBytes(tool, field, v.Int()) + } + } + + if maskDataURL != nil && strings.TrimSpace(*maskDataURL) != "" { + tool, _ = sjson.SetBytes(tool, "input_image_mask.image_url", strings.TrimSpace(*maskDataURL)) + } + + responsesReq := buildImagesResponsesRequest(prompt, images, tool) + if stream { + h.streamImagesFromResponses(c, responsesReq, responseFormat, "image_edit") + return + } + h.collectImagesFromResponses(c, responsesReq, responseFormat) +} + +func buildImagesResponsesRequest(prompt string, images []string, toolJSON []byte) []byte { + req := []byte(`{"instructions":"","stream":true,"reasoning":{"effort":"medium","summary":"auto"},"parallel_tool_calls":true,"include":["reasoning.encrypted_content"],"model":"","store":false,"tool_choice":{"type":"image_generation"}}`) + req, _ = sjson.SetBytes(req, "model", defaultImagesMainModel) + + input := []byte(`[{"type":"message","role":"user","content":[{"type":"input_text","text":""}]}]`) + input, _ = sjson.SetBytes(input, "0.content.0.text", prompt) + contentIndex := 1 + for _, img := range images { + if strings.TrimSpace(img) == "" { + continue + } + part := []byte(`{"type":"input_image","image_url":""}`) + part, _ = sjson.SetBytes(part, "image_url", img) + path := fmt.Sprintf("0.content.%d", contentIndex) + input, _ = sjson.SetRawBytes(input, path, part) + contentIndex++ + } + req, _ = sjson.SetRawBytes(req, "input", input) + + req, _ = sjson.SetRawBytes(req, "tools", []byte(`[]`)) + if len(toolJSON) > 0 && json.Valid(toolJSON) { + req, _ = sjson.SetRawBytes(req, "tools.-1", toolJSON) + } + return req +} + +func (h *OpenAIAPIHandler) collectImagesFromResponses(c *gin.Context, responsesReq []byte, responseFormat string) { + c.Header("Content-Type", "application/json") + + cliCtx, cliCancel := h.GetContextWithCancel(h, c, context.Background()) + stopKeepAlive := h.StartNonStreamingKeepAlive(c, cliCtx) + + dataChan, upstreamHeaders, errChan := h.ExecuteStreamWithAuthManager(cliCtx, "openai-response", defaultImagesMainModel, responsesReq, "") + + out, errMsg := collectImagesFromResponsesStream(cliCtx, dataChan, errChan, responseFormat) + stopKeepAlive() + if errMsg != nil { + h.WriteErrorResponse(c, errMsg) + if errMsg.Error != nil { + cliCancel(errMsg.Error) + } else { + cliCancel(nil) + } + return + } + handlers.WriteUpstreamHeaders(c.Writer.Header(), upstreamHeaders) + _, _ = c.Writer.Write(out) + cliCancel() +} + +func collectImagesFromResponsesStream(ctx context.Context, data <-chan []byte, errs <-chan *interfaces.ErrorMessage, responseFormat string) ([]byte, *interfaces.ErrorMessage) { + acc := &sseFrameAccumulator{} + + processFrame := func(frame []byte) ([]byte, bool, *interfaces.ErrorMessage) { + for _, line := range bytes.Split(frame, []byte("\n")) { + trimmed := bytes.TrimSpace(bytes.TrimRight(line, "\r")) + if len(trimmed) == 0 { + continue + } + if !bytes.HasPrefix(trimmed, []byte("data:")) { + continue + } + payload := bytes.TrimSpace(trimmed[len("data:"):]) + if len(payload) == 0 || bytes.Equal(payload, []byte("[DONE]")) { + continue + } + if !json.Valid(payload) { + return nil, false, &interfaces.ErrorMessage{StatusCode: http.StatusBadGateway, Error: fmt.Errorf("invalid SSE data JSON")} + } + + if gjson.GetBytes(payload, "type").String() != "response.completed" { + continue + } + + results, createdAt, usageRaw, firstMeta, err := extractImagesFromResponsesCompleted(payload) + if err != nil { + return nil, false, &interfaces.ErrorMessage{StatusCode: http.StatusBadGateway, Error: err} + } + if len(results) == 0 { + return nil, false, &interfaces.ErrorMessage{StatusCode: http.StatusBadGateway, Error: fmt.Errorf("upstream did not return image output")} + } + out, err := buildImagesAPIResponse(results, createdAt, usageRaw, firstMeta, responseFormat) + if err != nil { + return nil, false, &interfaces.ErrorMessage{StatusCode: http.StatusInternalServerError, Error: err} + } + return out, true, nil + } + return nil, false, nil + } + + for { + select { + case <-ctx.Done(): + return nil, &interfaces.ErrorMessage{StatusCode: http.StatusRequestTimeout, Error: ctx.Err()} + case errMsg, ok := <-errs: + if ok && errMsg != nil { + return nil, errMsg + } + errs = nil + case chunk, ok := <-data: + if !ok { + for _, frame := range acc.Flush() { + if out, done, errMsg := processFrame(frame); errMsg != nil { + return nil, errMsg + } else if done { + return out, nil + } + } + return nil, &interfaces.ErrorMessage{StatusCode: http.StatusBadGateway, Error: fmt.Errorf("stream disconnected before completion")} + } + for _, frame := range acc.AddChunk(chunk) { + if out, done, errMsg := processFrame(frame); errMsg != nil { + return nil, errMsg + } else if done { + return out, nil + } + } + } + } +} + +func extractImagesFromResponsesCompleted(payload []byte) (results []imageCallResult, createdAt int64, usageRaw []byte, firstMeta imageCallResult, err error) { + if gjson.GetBytes(payload, "type").String() != "response.completed" { + return nil, 0, nil, imageCallResult{}, fmt.Errorf("unexpected event type") + } + + createdAt = gjson.GetBytes(payload, "response.created_at").Int() + if createdAt <= 0 { + createdAt = time.Now().Unix() + } + + output := gjson.GetBytes(payload, "response.output") + if output.IsArray() { + for _, item := range output.Array() { + if item.Get("type").String() != "image_generation_call" { + continue + } + res := strings.TrimSpace(item.Get("result").String()) + if res == "" { + continue + } + entry := imageCallResult{ + Result: res, + RevisedPrompt: strings.TrimSpace(item.Get("revised_prompt").String()), + OutputFormat: strings.TrimSpace(item.Get("output_format").String()), + Size: strings.TrimSpace(item.Get("size").String()), + Background: strings.TrimSpace(item.Get("background").String()), + Quality: strings.TrimSpace(item.Get("quality").String()), + } + if len(results) == 0 { + firstMeta = entry + } + results = append(results, entry) + } + } + + if usage := gjson.GetBytes(payload, "response.tool_usage.image_gen"); usage.Exists() && usage.IsObject() { + usageRaw = []byte(usage.Raw) + } + + return results, createdAt, usageRaw, firstMeta, nil +} + +func buildImagesAPIResponse(results []imageCallResult, createdAt int64, usageRaw []byte, firstMeta imageCallResult, responseFormat string) ([]byte, error) { + out := []byte(`{"created":0,"data":[]}`) + out, _ = sjson.SetBytes(out, "created", createdAt) + + responseFormat = strings.ToLower(strings.TrimSpace(responseFormat)) + if responseFormat == "" { + responseFormat = "b64_json" + } + + for _, img := range results { + item := []byte(`{}`) + if responseFormat == "url" { + mt := mimeTypeFromOutputFormat(img.OutputFormat) + item, _ = sjson.SetBytes(item, "url", "data:"+mt+";base64,"+img.Result) + } else { + item, _ = sjson.SetBytes(item, "b64_json", img.Result) + } + if img.RevisedPrompt != "" { + item, _ = sjson.SetBytes(item, "revised_prompt", img.RevisedPrompt) + } + out, _ = sjson.SetRawBytes(out, "data.-1", item) + } + + if firstMeta.Background != "" { + out, _ = sjson.SetBytes(out, "background", firstMeta.Background) + } + if firstMeta.OutputFormat != "" { + out, _ = sjson.SetBytes(out, "output_format", firstMeta.OutputFormat) + } + if firstMeta.Quality != "" { + out, _ = sjson.SetBytes(out, "quality", firstMeta.Quality) + } + if firstMeta.Size != "" { + out, _ = sjson.SetBytes(out, "size", firstMeta.Size) + } + + if len(usageRaw) > 0 && json.Valid(usageRaw) { + out, _ = sjson.SetRawBytes(out, "usage", usageRaw) + } + + return out, nil +} + +func (h *OpenAIAPIHandler) streamImagesFromResponses(c *gin.Context, responsesReq []byte, responseFormat string, streamPrefix string) { + flusher, ok := c.Writer.(http.Flusher) + if !ok { + c.JSON(http.StatusInternalServerError, handlers.ErrorResponse{ + Error: handlers.ErrorDetail{ + Message: "Streaming not supported", + Type: "server_error", + }, + }) + return + } + + cliCtx, cliCancel := h.GetContextWithCancel(h, c, context.Background()) + dataChan, upstreamHeaders, errChan := h.ExecuteStreamWithAuthManager(cliCtx, "openai-response", defaultImagesMainModel, responsesReq, "") + + setSSEHeaders := func() { + c.Header("Content-Type", "text/event-stream") + c.Header("Cache-Control", "no-cache") + c.Header("Connection", "keep-alive") + c.Header("Access-Control-Allow-Origin", "*") + } + + writeEvent := func(eventName string, dataJSON []byte) { + if strings.TrimSpace(eventName) != "" { + _, _ = fmt.Fprintf(c.Writer, "event: %s\n", eventName) + } + _, _ = fmt.Fprintf(c.Writer, "data: %s\n\n", string(dataJSON)) + flusher.Flush() + } + + // Peek for first chunk/error so we can still return a JSON error body. + for { + select { + case <-c.Request.Context().Done(): + cliCancel(c.Request.Context().Err()) + return + case errMsg, ok := <-errChan: + if !ok { + errChan = nil + continue + } + h.WriteErrorResponse(c, errMsg) + if errMsg != nil { + cliCancel(errMsg.Error) + } else { + cliCancel(nil) + } + return + case chunk, ok := <-dataChan: + if !ok { + setSSEHeaders() + handlers.WriteUpstreamHeaders(c.Writer.Header(), upstreamHeaders) + _, _ = c.Writer.Write([]byte("\n")) + flusher.Flush() + cliCancel(nil) + return + } + + setSSEHeaders() + handlers.WriteUpstreamHeaders(c.Writer.Header(), upstreamHeaders) + + h.forwardImagesStream(cliCtx, c, flusher, func(err error) { cliCancel(err) }, dataChan, errChan, chunk, responseFormat, streamPrefix, writeEvent) + return + } + } +} + +func (h *OpenAIAPIHandler) forwardImagesStream(ctx context.Context, c *gin.Context, flusher http.Flusher, cancel func(error), data <-chan []byte, errs <-chan *interfaces.ErrorMessage, firstChunk []byte, responseFormat string, streamPrefix string, writeEvent func(string, []byte)) { + acc := &sseFrameAccumulator{} + + responseFormat = strings.ToLower(strings.TrimSpace(responseFormat)) + if responseFormat == "" { + responseFormat = "b64_json" + } + + emitError := func(errMsg *interfaces.ErrorMessage) { + if errMsg == nil { + return + } + status := http.StatusInternalServerError + if errMsg.StatusCode > 0 { + status = errMsg.StatusCode + } + errText := http.StatusText(status) + if errMsg.Error != nil && strings.TrimSpace(errMsg.Error.Error()) != "" { + errText = errMsg.Error.Error() + } + body := handlers.BuildErrorResponseBody(status, errText) + writeEvent("error", body) + } + + processFrame := func(frame []byte) (done bool) { + for _, line := range bytes.Split(frame, []byte("\n")) { + trimmed := bytes.TrimSpace(bytes.TrimRight(line, "\r")) + if len(trimmed) == 0 || !bytes.HasPrefix(trimmed, []byte("data:")) { + continue + } + payload := bytes.TrimSpace(trimmed[len("data:"):]) + if len(payload) == 0 || bytes.Equal(payload, []byte("[DONE]")) || !json.Valid(payload) { + continue + } + + switch gjson.GetBytes(payload, "type").String() { + case "response.image_generation_call.partial_image": + b64 := strings.TrimSpace(gjson.GetBytes(payload, "partial_image_b64").String()) + if b64 == "" { + continue + } + outputFormat := strings.TrimSpace(gjson.GetBytes(payload, "output_format").String()) + index := gjson.GetBytes(payload, "partial_image_index").Int() + eventName := streamPrefix + ".partial_image" + data := []byte(`{"type":"","partial_image_index":0}`) + data, _ = sjson.SetBytes(data, "type", eventName) + data, _ = sjson.SetBytes(data, "partial_image_index", index) + if responseFormat == "url" { + mt := mimeTypeFromOutputFormat(outputFormat) + data, _ = sjson.SetBytes(data, "url", "data:"+mt+";base64,"+b64) + } else { + data, _ = sjson.SetBytes(data, "b64_json", b64) + } + writeEvent(eventName, data) + case "response.completed": + results, _, usageRaw, _, err := extractImagesFromResponsesCompleted(payload) + if err != nil { + emitError(&interfaces.ErrorMessage{StatusCode: http.StatusBadGateway, Error: err}) + return true + } + if len(results) == 0 { + emitError(&interfaces.ErrorMessage{StatusCode: http.StatusBadGateway, Error: fmt.Errorf("upstream did not return image output")}) + return true + } + eventName := streamPrefix + ".completed" + for _, img := range results { + data := []byte(`{"type":""}`) + data, _ = sjson.SetBytes(data, "type", eventName) + if responseFormat == "url" { + mt := mimeTypeFromOutputFormat(img.OutputFormat) + data, _ = sjson.SetBytes(data, "url", "data:"+mt+";base64,"+img.Result) + } else { + data, _ = sjson.SetBytes(data, "b64_json", img.Result) + } + if len(usageRaw) > 0 && json.Valid(usageRaw) { + data, _ = sjson.SetRawBytes(data, "usage", usageRaw) + } + writeEvent(eventName, data) + } + return true + } + } + return false + } + + for _, frame := range acc.AddChunk(firstChunk) { + if processFrame(frame) { + cancel(nil) + return + } + } + + for { + select { + case <-c.Request.Context().Done(): + cancel(c.Request.Context().Err()) + return + case errMsg, ok := <-errs: + if ok && errMsg != nil { + emitError(errMsg) + cancel(errMsg.Error) + return + } + errs = nil + case chunk, ok := <-data: + if !ok { + for _, frame := range acc.Flush() { + if processFrame(frame) { + cancel(nil) + return + } + } + cancel(nil) + return + } + for _, frame := range acc.AddChunk(chunk) { + if processFrame(frame) { + cancel(nil) + return + } + } + } + } +} diff --git a/sdk/cliproxy/auth/conductor.go b/sdk/cliproxy/auth/conductor.go index 6628c20b8a..4265c93f17 100644 --- a/sdk/cliproxy/auth/conductor.go +++ b/sdk/cliproxy/auth/conductor.go @@ -63,9 +63,14 @@ const ( refreshCheckInterval = 5 * time.Second refreshMaxConcurrency = 16 refreshPendingBackoff = time.Minute - refreshFailureBackoff = 1 * time.Minute - quotaBackoffBase = time.Second - quotaBackoffMax = 30 * time.Minute + refreshFailureBackoff = 5 * time.Minute + // refreshIneffectiveBackoff throttles refresh attempts when an executor returns + // success but the auth still evaluates as needing refresh (e.g. token expiry + // wasn't updated). Without this guard, the auto-refresh loop can tight-loop and + // burn CPU at idle. + refreshIneffectiveBackoff = 30 * time.Second + quotaBackoffBase = time.Second + quotaBackoffMax = 30 * time.Minute ) var quotaCooldownDisabled atomic.Bool @@ -3242,6 +3247,9 @@ func (m *Manager) refreshAuth(ctx context.Context, id string) { // If the Authenticator did not set it (zero value), shouldRefresh will use default logic updated.LastError = nil updated.UpdatedAt = now + if m.shouldRefresh(updated, now) { + updated.NextRefreshAfter = now.Add(refreshIneffectiveBackoff) + } _, _ = m.Update(ctx, updated) } diff --git a/sdk/cliproxy/service.go b/sdk/cliproxy/service.go index 0079df73ee..30f2b7900f 100644 --- a/sdk/cliproxy/service.go +++ b/sdk/cliproxy/service.go @@ -1477,7 +1477,7 @@ func buildCodexConfigModels(entry *config.CodexKey) []*ModelInfo { if entry == nil { return nil } - return buildConfigModels(entry.Models, "openai", "openai") + return registry.WithCodexBuiltins(buildConfigModels(entry.Models, "openai", "openai")) } func rewriteModelInfoName(name, oldID, newID string) string { diff --git a/test/thinking_conversion_test.go b/test/thinking_conversion_test.go index 984d35641a..f5980b6dc2 100644 --- a/test/thinking_conversion_test.go +++ b/test/thinking_conversion_test.go @@ -2,7 +2,6 @@ package test import ( "fmt" - "strings" "testing" "time" @@ -1066,12 +1065,12 @@ func TestThinkingE2EMatrix_Suffix(t *testing.T) { expectErr: false, }, - // Gemini Family Cross-Channel Consistency (Cases 106-114) + // Gemini Family Cross-Channel Consistency (Cases 90-95) // Tests that gemini/gemini-cli/antigravity as same API family should have consistent validation behavior - // Case 106: Gemini to Antigravity, budget 64000 (suffix) → clamped to Max + // Case 90: Gemini to Antigravity, budget 64000 (suffix) → clamped to Max { - name: "106", + name: "90", from: "gemini", to: "antigravity", model: "gemini-budget-model(64000)", @@ -1081,9 +1080,9 @@ func TestThinkingE2EMatrix_Suffix(t *testing.T) { includeThoughts: "true", expectErr: false, }, - // Case 107: Gemini to Gemini-CLI, budget 64000 (suffix) → clamped to Max + // Case 91: Gemini to Gemini-CLI, budget 64000 (suffix) → clamped to Max { - name: "107", + name: "91", from: "gemini", to: "gemini-cli", model: "gemini-budget-model(64000)", @@ -1093,9 +1092,9 @@ func TestThinkingE2EMatrix_Suffix(t *testing.T) { includeThoughts: "true", expectErr: false, }, - // Case 108: Gemini-CLI to Antigravity, budget 64000 (suffix) → clamped to Max + // Case 92: Gemini-CLI to Antigravity, budget 64000 (suffix) → clamped to Max { - name: "108", + name: "92", from: "gemini-cli", to: "antigravity", model: "gemini-budget-model(64000)", @@ -1105,9 +1104,9 @@ func TestThinkingE2EMatrix_Suffix(t *testing.T) { includeThoughts: "true", expectErr: false, }, - // Case 109: Gemini-CLI to Gemini, budget 64000 (suffix) → clamped to Max + // Case 93: Gemini-CLI to Gemini, budget 64000 (suffix) → clamped to Max { - name: "109", + name: "93", from: "gemini-cli", to: "gemini", model: "gemini-budget-model(64000)", @@ -1117,9 +1116,9 @@ func TestThinkingE2EMatrix_Suffix(t *testing.T) { includeThoughts: "true", expectErr: false, }, - // Case 110: Gemini to Antigravity, budget 8192 → passthrough (normal value) + // Case 94: Gemini to Antigravity, budget 8192 → passthrough (normal value) { - name: "110", + name: "94", from: "gemini", to: "antigravity", model: "gemini-budget-model(8192)", @@ -1129,9 +1128,9 @@ func TestThinkingE2EMatrix_Suffix(t *testing.T) { includeThoughts: "true", expectErr: false, }, - // Case 111: Gemini-CLI to Antigravity, budget 8192 → passthrough (normal value) + // Case 95: Gemini-CLI to Antigravity, budget 8192 → passthrough (normal value) { - name: "111", + name: "95", from: "gemini-cli", to: "antigravity", model: "gemini-budget-model(8192)", @@ -2283,12 +2282,12 @@ func TestThinkingE2EMatrix_Body(t *testing.T) { expectErr: true, }, - // Gemini Family Cross-Channel Consistency (Cases 106-114) + // Gemini Family Cross-Channel Consistency (Cases 90-95) // Tests that gemini/gemini-cli/antigravity as same API family should have consistent validation behavior - // Case 106: Gemini to Antigravity, thinkingBudget=64000 → exceeds Max error (same family strict validation) + // Case 90: Gemini to Antigravity, thinkingBudget=64000 → exceeds Max error (same family strict validation) { - name: "106", + name: "90", from: "gemini", to: "antigravity", model: "gemini-budget-model", @@ -2296,9 +2295,9 @@ func TestThinkingE2EMatrix_Body(t *testing.T) { expectField: "", expectErr: true, }, - // Case 107: Gemini to Gemini-CLI, thinkingBudget=64000 → exceeds Max error (same family strict validation) + // Case 91: Gemini to Gemini-CLI, thinkingBudget=64000 → exceeds Max error (same family strict validation) { - name: "107", + name: "91", from: "gemini", to: "gemini-cli", model: "gemini-budget-model", @@ -2306,9 +2305,9 @@ func TestThinkingE2EMatrix_Body(t *testing.T) { expectField: "", expectErr: true, }, - // Case 108: Gemini-CLI to Antigravity, thinkingBudget=64000 → exceeds Max error (same family strict validation) + // Case 92: Gemini-CLI to Antigravity, thinkingBudget=64000 → exceeds Max error (same family strict validation) { - name: "108", + name: "92", from: "gemini-cli", to: "antigravity", model: "gemini-budget-model", @@ -2316,9 +2315,9 @@ func TestThinkingE2EMatrix_Body(t *testing.T) { expectField: "", expectErr: true, }, - // Case 109: Gemini-CLI to Gemini, thinkingBudget=64000 → exceeds Max error (same family strict validation) + // Case 93: Gemini-CLI to Gemini, thinkingBudget=64000 → exceeds Max error (same family strict validation) { - name: "109", + name: "93", from: "gemini-cli", to: "gemini", model: "gemini-budget-model", @@ -2326,9 +2325,9 @@ func TestThinkingE2EMatrix_Body(t *testing.T) { expectField: "", expectErr: true, }, - // Case 110: Gemini to Antigravity, thinkingBudget=8192 → passthrough (normal value) + // Case 94: Gemini to Antigravity, thinkingBudget=8192 → passthrough (normal value) { - name: "110", + name: "94", from: "gemini", to: "antigravity", model: "gemini-budget-model", @@ -2338,9 +2337,9 @@ func TestThinkingE2EMatrix_Body(t *testing.T) { includeThoughts: "true", expectErr: false, }, - // Case 111: Gemini-CLI to Antigravity, thinkingBudget=8192 → passthrough (normal value) + // Case 95: Gemini-CLI to Antigravity, thinkingBudget=8192 → passthrough (normal value) { - name: "111", + name: "95", from: "gemini-cli", to: "antigravity", model: "gemini-budget-model",