n9router 0.4.44 → 0.4.46

package/app/open-sse/providers/registry/venice.js

@@ -0,0 +1,56 @@
+export default {
+  id: "venice",
+  priority: 115,
+  alias: "venice",
+  aliases: [
+    "vn",
+  ],
+  uiAlias: "venice",
+  display: {
+    name: "Venice AI",
+    icon: "shield",
+    color: "#DC2626",
+    textIcon: "VE",
+    website: "https://venice.ai",
+    notice: {
+      text: "OpenAI-compatible. Private inference + uncensored models (Venice Uncensored, GLM, Qwen, DeepSeek, Llama).",
+      apiKeyUrl: "https://venice.ai/settings/api",
+    },
+  },
+  category: "apikey",
+  transport: {
+    baseUrl: "https://api.venice.ai/api/v1/chat/completions",
+    validateUrl: "https://api.venice.ai/api/v1/models",
+    thinkingFormat: "openai",
+  },
+  // Curated seed; the full live catalogue (90+ text models) is fetched via
+  // modelsFetcher and any other id is accepted via passthroughModels.
+  models: [
+    { id: "venice-uncensored-1-2", name: "Venice Uncensored 1.2" },
+    { id: "zai-org-glm-5", name: "GLM-5" },
+    { id: "qwen3-235b-a22b-instruct-2507", name: "Qwen3 235B A22B Instruct" },
+    { id: "qwen3-coder-480b-a35b-instruct-turbo", name: "Qwen3 Coder 480B A35B Turbo" },
+    { id: "qwen3-vl-235b-a22b", name: "Qwen3 VL 235B A22B" },
+    { id: "deepseek-v4-pro", name: "DeepSeek V4 Pro" },
+    { id: "llama-3.3-70b", name: "Llama 3.3 70B" },
+    { id: "hermes-3-llama-3.1-405b", name: "Hermes 3 Llama 3.1 405B" },
+    { id: "mistral-small-3-2-24b-instruct", name: "Mistral Small 3.2 24B" },
+    { id: "text-embedding-3-large", name: "Text Embedding 3 Large", kind: "embedding" },
+    { id: "text-embedding-bge-m3", name: "BGE-M3 Embedding", kind: "embedding" },
+    { id: "text-embedding-qwen3-8b", name: "Qwen3 8B Embedding", kind: "embedding" },
+    { id: "venice-sd35", name: "Venice SD3.5", params: ["n", "size"], kind: "image" },
+    { id: "flux-2-pro", name: "FLUX.2 Pro", params: ["n", "size"], kind: "image" },
+    { id: "gpt-image-2", name: "GPT Image 2 (via Venice)", params: ["n", "size", "quality"], kind: "image" },
+  ],
+  serviceKinds: ["llm", "embedding", "image"],
+  embeddingConfig: {
+    baseUrl: "https://api.venice.ai/api/v1/embeddings",
+    authType: "apikey",
+    authHeader: "bearer",
+  },
+  imageConfig: {
+    baseUrl: "https://api.venice.ai/api/v1/images/generations",
+  },
+  modelsFetcher: { url: "https://api.venice.ai/api/v1/models", type: "openai" },
+  passthroughModels: true,
+};

package/app/open-sse/providers/registry/vercel-ai-gateway.js

@@ -0,0 +1,41 @@
+export default {
+  id: "vercel-ai-gateway",
+  priority: 160,
+  alias: "vercel-ai-gateway",
+  aliases: [
+    "vercel",
+  ],
+  uiAlias: "vercel",
+  display: {
+    name: "Vercel AI Gateway",
+    icon: "deployed_code",
+    color: "#111827",
+    textIcon: "VG",
+    website: "https://vercel.com/ai-gateway",
+    notice: {
+      text: "Unified OpenAI-compatible endpoint from Vercel. Use your AI Gateway API key, then pick models with provider/model IDs like anthropic/claude-sonnet-4.6 or openai/gpt-5.4.",
+      apiKeyUrl: "https://vercel.com/dashboard/~/ai-gateway",
+    },
+  },
+  category: "apikey",
+  transport: {
+    baseUrl: "https://ai-gateway.vercel.sh/v1/chat/completions",
+    thinkingFormat: "openai",
+    retry: {
+      "429": 2,
+    },
+    usage: {
+      url: "https://ai-gateway.vercel.sh/v1/credits",
+    },
+  },
+  serviceKinds: ["llm","embedding","image","imageToText","webSearch"],
+  embeddingConfig: { baseUrl: "https://ai-gateway.vercel.sh/v1/embeddings" },
+  imageConfig: { baseUrl: "https://ai-gateway.vercel.sh/v1/images/generations" },
+  searchViaChat: { defaultModel: "openai/gpt-4o-mini", pricingUrl: "https://vercel.com/docs/ai-gateway/pricing" },
+  modelsFetcher: { url: "https://ai-gateway.vercel.sh/v1/models", type: "openai" },
+  passthroughModels: true,
+  features: {
+    usage: true,
+    usageApikey: true,
+  },
+};

package/app/open-sse/providers/registry/vertex-partner.js

@@ -0,0 +1,29 @@
+export default {
+  id: "vertex-partner",
+  priority: 260,
+  alias: "vertex-partner",
+  aliases: [
+    "vxp",
+  ],
+  uiAlias: "vxp",
+  display: {
+    name: "Vertex Partner",
+    icon: "cloud",
+    color: "#34A853",
+    textIcon: "VP",
+    website: "https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-partner-models",
+    notice: {
+      apiKeyUrl: "https://console.cloud.google.com/iam-admin/serviceaccounts",
+    },
+  },
+  category: "apikey",
+  transport: {
+    baseUrl: "https://aiplatform.googleapis.com",
+  },
+  models: [
+    { id: "deepseek-ai/deepseek-v3.2-maas", name: "DeepSeek V3.2 (Vertex)" },
+    { id: "qwen/qwen3-next-80b-a3b-thinking-maas", name: "Qwen3 Next 80B Thinking (Vertex)" },
+    { id: "qwen/qwen3-next-80b-a3b-instruct-maas", name: "Qwen3 Next 80B Instruct (Vertex)" },
+    { id: "zai-org/glm-5-maas", name: "GLM-5 (Vertex)" },
+  ],
+};

package/app/open-sse/providers/registry/vertex.js

@@ -0,0 +1,32 @@
+export default {
+  id: "vertex",
+  priority: 40,
+  alias: "vertex",
+  aliases: [
+    "vx",
+  ],
+  uiAlias: "vx",
+  display: {
+    name: "Vertex AI",
+    icon: "cloud",
+    color: "#4285F4",
+    textIcon: "VX",
+    website: "https://cloud.google.com/vertex-ai",
+    notice: {
+      text: "New Google Cloud accounts get $300 free credits. Requires GCP project + Service Account with Vertex AI API enabled.",
+      apiKeyUrl: "https://console.cloud.google.com/iam-admin/serviceaccounts",
+    },
+  },
+  category: "freeTier",
+  transport: {
+    baseUrl: "https://aiplatform.googleapis.com",
+    format: "vertex",
+  },
+  models: [
+    { id: "gemini-3.1-pro-preview", name: "Gemini 3.1 Pro Preview" },
+    { id: "gemini-3.1-flash-lite-preview", name: "Gemini 3.1 Flash Lite Preview" },
+    { id: "gemini-3-flash-preview", name: "Gemini 3 Flash Preview" },
+    { id: "gemini-2.5-flash", name: "Gemini 2.5 Flash" },
+  ],
+  serviceKinds: ["llm","imageToText"],
+};

package/app/open-sse/providers/registry/volcengine-ark.js

@@ -0,0 +1,35 @@
+export default {
+  id: "volcengine-ark",
+  priority: 270,
+  alias: "volcengine-ark",
+  aliases: [
+    "ark",
+  ],
+  uiAlias: "ark",
+  display: {
+    name: "Volcengine Ark",
+    icon: "cloud",
+    color: "#1677FF",
+    textIcon: "ARK",
+    website: "https://ark.cn-beijing.volces.com",
+    notice: {
+      apiKeyUrl: "https://console.volcengine.com/ark/region:ark+cn-beijing/apiKey",
+    },
+  },
+  category: "apikey",
+  transport: {
+    baseUrl: "https://ark.cn-beijing.volces.com/api/coding/v3/chat/completions",
+    headers: {},
+  },
+  models: [
+    { id: "Doubao-Seed-2.0-Code", name: "Doubao-Seed-2.0-Code" },
+    { id: "Doubao-Seed-2.0-pro", name: "Doubao-Seed-2.0-pro" },
+    { id: "Doubao-Seed-2.0-lite", name: "Doubao-Seed-2.0-lite" },
+    { id: "Doubao-Seed-Code", name: "Doubao-Seed-Code" },
+    { id: "DeepSeek-V4-Flash", name: "DeepSeek-V4-Flash" },
+    { id: "DeepSeek-V4-Pro", name: "DeepSeek-V4-Pro" },
+    { id: "GLM-5.1", name: "GLM-5.1" },
+    { id: "MiniMax-M2.7", name: "MiniMax-M2.7" },
+    { id: "Kimi-K2.6", name: "Kimi-K2.6" },
+  ],
+};

package/app/open-sse/providers/registry/voyage-ai.js

@@ -0,0 +1,30 @@
+export default {
+  id: "voyage-ai",
+  priority: 40,
+  alias: "voyage-ai",
+  uiAlias: "voyage",
+  display: {
+    name: "Voyage AI",
+    icon: "data_array",
+    color: "#0EA5E9",
+    textIcon: "VG",
+    website: "https://www.voyageai.com",
+    notice: {
+      apiKeyUrl: "https://dash.voyageai.com/api-keys",
+    },
+  },
+  category: "apikey",
+  authType: "apikey",
+  transport: null,
+  models: [
+    { id: "voyage-3-large", name: "Voyage 3 Large", kind: "embedding" },
+    { id: "voyage-3.5", name: "Voyage 3.5", kind: "embedding" },
+    { id: "voyage-3.5-lite", name: "Voyage 3.5 Lite", kind: "embedding" },
+    { id: "voyage-code-3", name: "Voyage Code 3", kind: "embedding" },
+    { id: "voyage-finance-2", name: "Voyage Finance 2", kind: "embedding" },
+    { id: "voyage-law-2", name: "Voyage Law 2", kind: "embedding" },
+    { id: "voyage-multilingual-2", name: "Voyage Multilingual 2", kind: "embedding" },
+  ],
+  serviceKinds: ["embedding"],
+  embeddingConfig: { baseUrl: "https://api.voyageai.com/v1/embeddings" },
+};

package/app/open-sse/providers/registry/xai.js

@@ -0,0 +1,43 @@
+export default {
+  id: "xai",
+  priority: 280,
+  alias: "xai",
+  display: {
+    name: "xAI (Grok)",
+    icon: "auto_awesome",
+    color: "#1DA1F2",
+    textIcon: "XA",
+    website: "https://x.ai",
+    notice: {
+      apiKeyUrl: "https://console.x.ai",
+    },
+  },
+  category: "oauth",
+  authModes: [
+    "oauth",
+    "apikey",
+  ],
+  hasOAuth: true,
+  transport: {
+    baseUrl: "https://api.x.ai/v1/chat/completions",
+    validateUrl: "https://api.x.ai/v1/models",
+    responsesUrl: "https://api.x.ai/v1/responses",
+    clientId: "b1a00492-073a-47ea-816f-4c329264a828",
+    tokenUrl: "https://auth.x.ai/oauth2/token",
+    refreshUrl: "https://auth.x.ai/oauth2/token",
+  },
+  models: [
+    { id: "grok-4", name: "Grok 4" },
+    { id: "grok-4-fast-reasoning", name: "Grok 4 Fast Reasoning" },
+    { id: "grok-code-fast-1", name: "Grok Code Fast" },
+    { id: "grok-3", name: "Grok 3" },
+    { id: "grok-2-image-1212", name: "Grok 2 Image", params: ["n","response_format"], kind: "image" },
+  ],
+  serviceKinds: ["llm","imageToText","webSearch","image"],
+  imageConfig: { baseUrl: "https://api.x.ai/v1/images/generations", bodyFields: ["model","prompt","n","response_format"] },
+  searchViaChat: {
+    defaultModel: "grok-4.20-reasoning",
+    endpoint: "https://api.x.ai/v1/responses",
+    pricingUrl: "https://x.ai/api#pricing",
+  },
+};

package/app/open-sse/providers/registry/xiaomi-mimo.js

@@ -0,0 +1,46 @@
+import { CLAUDE_API_HEADERS } from "../shared.js";
+
+export default {
+  id: "xiaomi-mimo",
+  priority: 290,
+  alias: "xiaomi-mimo",
+  aliases: [
+    "mimo",
+  ],
+  uiAlias: "mimo",
+  display: {
+    name: "Xiaomi MiMo",
+    icon: "smart_toy",
+    color: "#FF6900",
+    textIcon: "XM",
+    website: "https://xiaomimimo.com",
+    notice: {
+      apiKeyUrl: "https://xiaomimimo.com",
+    },
+  },
+  category: "apikey",
+  transport: {
+    baseUrl: "https://api.xiaomimimo.com/v1/chat/completions",
+    validateUrl: "https://api.xiaomimimo.com/v1/models",
+  },
+  // Multi-endpoint: pick the transport matching client sourceFormat to skip translation.
+  transports: [
+    {
+      format: "openai",
+      baseUrl: "https://api.xiaomimimo.com/v1/chat/completions",
+      auth: { combined: true, header: "Authorization", scheme: "bearer" },
+    },
+    {
+      format: "claude",
+      baseUrl: "https://api.xiaomimimo.com/anthropic/v1/messages",
+      headers: { ...CLAUDE_API_HEADERS },
+      auth: { combined: true, header: "x-api-key", scheme: "raw" },
+    },
+  ],
+  models: [
+    { id: "mimo-v2.5-pro", name: "MiMo V2.5 Pro" },
+    { id: "mimo-v2.5", name: "MiMo V2.5" },
+    { id: "mimo-v2-omni", name: "MiMo V2 Omni" },
+    { id: "mimo-v2-flash", name: "MiMo V2 Flash" },
+  ],
+};

package/app/open-sse/providers/registry/xiaomi-tokenplan.js

@@ -0,0 +1,63 @@
+import { CLAUDE_API_HEADERS } from "../shared.js";
+
+export default {
+  id: "xiaomi-tokenplan",
+  priority: 300,
+  alias: "xiaomi-tokenplan",
+  aliases: [
+    "xmtp",
+  ],
+  uiAlias: "xmtp",
+  display: {
+    name: "Xiaomi MiMo (Token Plan)",
+    icon: "smart_toy",
+    color: "#FF6700",
+    textIcon: "XT",
+    website: "https://mimo.xiaomi.com",
+    notice: {
+      text: "Xiaomi MiMo Token Plan subscription (API key starts with tp-). Token Plan keys are cluster-specific — select the region matching your subscription.",
+      apiKeyUrl: "https://mimo.xiaomi.com",
+    },
+  },
+  category: "apikey",
+  hasProviderSpecificData: true,
+  regions: [
+    { id: "sgp", label: "Singapore (新加坡)" },
+    { id: "cn", label: "China (中国大陆)" },
+    { id: "ams", label: "Amsterdam (阿姆斯特丹)" },
+  ],
+  defaultRegion: "sgp",
+  transport: {
+    baseUrl: "https://token-plan-sgp.xiaomimimo.com/v1/chat/completions",
+    regions: {
+      sgp: "https://token-plan-sgp.xiaomimimo.com/v1",
+      cn: "https://token-plan-cn.xiaomimimo.com/v1",
+      ams: "https://token-plan-ams.xiaomimimo.com/v1",
+    },
+    defaultRegion: "sgp",
+  },
+  // Multi-endpoint: pick the transport matching client sourceFormat to skip translation.
+  // baseUrl omitted — region-dynamic, resolved in the executor's buildUrl.
+  transports: [
+    {
+      format: "openai",
+      auth: { combined: true, header: "Authorization", scheme: "bearer" },
+    },
+    {
+      format: "claude",
+      headers: { ...CLAUDE_API_HEADERS },
+      auth: { combined: true, header: "x-api-key", scheme: "raw" },
+    },
+  ],
+  models: [
+    { id: "mimo-v2.5-pro", name: "MiMo V2.5 Pro" },
+    { id: "mimo-v2.5-pro-claude", name: "MiMo V2.5 Pro (Claude Native)", targetFormat: "claude", upstreamModelId: "mimo-v2.5-pro" },
+    { id: "mimo-v2.5", name: "MiMo V2.5" },
+    { id: "mimo-v2-pro", name: "MiMo V2 Pro" },
+    { id: "mimo-v2-omni", name: "MiMo V2 Omni" },
+    { id: "mimo-v2-tts", name: "MiMo V2 TTS" },
+    { id: "mimo-v2.5-tts", name: "MiMo V2.5 TTS" },
+    { id: "mimo-v2.5-tts-voiceclone", name: "MiMo V2.5 TTS Voice Clone" },
+    { id: "mimo-v2.5-tts-voicedesign", name: "MiMo V2.5 TTS Voice Design" },
+  ],
+};

package/app/open-sse/providers/registry/youcom.js

@@ -0,0 +1,35 @@
+export default {
+  id: "youcom",
+  alias: "youcom",
+  display: {
+    name: "You.com Search",
+    icon: "search",
+    color: "#7C3AED",
+    textIcon: "YC",
+    website: "https://you.com",
+    notice: {
+      apiKeyUrl: "https://api.you.com"
+    }
+  },
+  category: "apikey",
+  authType: "apikey",
+  serviceKinds: [
+    "webSearch"
+  ],
+  searchConfig: {
+    baseUrl: "https://ydc-index.io/v1/search",
+    method: "GET",
+    authType: "apikey",
+    authHeader: "x-api-key",
+    costPerQuery: 0.005,
+    freeMonthlyQuota: 0,
+    searchTypes: [
+      "web",
+      "news"
+    ],
+    defaultMaxResults: 5,
+    maxMaxResults: 100,
+    timeoutMs: 10000,
+    cacheTTLMs: 300000
+  }
+};

package/app/open-sse/providers/schema.js

@@ -0,0 +1,76 @@
+// Provider transport schema: shared defaults + endpoint defaults + resolver (skeleton, not wired)
+import { DEFAULT_RETRY_CONFIG, FETCH_CONNECT_TIMEOUT_MS } from "../config/runtimeConfig.js";
+
+/**
+ * RegistryEntry shape — full contract for registry/{id}.js. See REGISTRY_TEMPLATE.js for a worked example.
+ * Only `id` + `category` are strictly required; everything else is optional/derived.
+ *
+ * @typedef {Object} RegistryEntry
+ * @property {string}   id            Unique provider id (kebab-case). REQUIRED.
+ * @property {string}  [alias]        Short key for PROVIDER_MODELS (defaults to id).
+ * @property {string[]}[aliases]      Extra lookup tokens resolving to this provider.
+ * @property {string}  [uiAlias]      Token shown in UI badges.
+ * @property {string}   category      "apikey"|"oauth"|"freeTier"|... drives UI grouping. REQUIRED.
+ * @property {string}  [authType]     "apikey"|"oauth" auth hint.
+ * @property {string[]}[authModes]    Allowed auth modes when provider supports both.
+ * @property {boolean} [hasOAuth]     Provider exposes an OAuth flow.
+ * @property {boolean} [noAuth]       Provider needs no credentials (local/free).
+ * @property {Object}  [display]      UI: {name,icon,color,textIcon,website,notice,deprecated,deprecationNotice,kindNotice,mediaPriority}.
+ * @property {Object}  [transport]    Runtime HTTP config (see TransportConfig below). Builds PROVIDERS[id].
+ * @property {Object}  [oauth]        OAuth flow config (see OAuthConfig). Builds PROVIDER_OAUTH[id].
+ * @property {Object}  [media]        Non-LLM services (see MediaConfig). Builds PROVIDER_MEDIA[id].
+ * @property {Array}   [models]       Model list; omit = no model key, [] = explicit empty.
+ * @property {Object}  [features]     Feature flags, e.g. {usage:true}.
+ * @property {Object}  [thinkingConfig] Reasoning UI: {options:[...],defaultMode}.
+ * @property {boolean} [passthroughModels] Forward client model id untouched.
+ *
+ * TransportConfig: { baseUrl, format, headers, auth, forceStream, urlSuffix, quirks, retry, timeoutMs,
+ *   executor, clientId, clientSecret, tokenUrl, refreshUrl, usage, cliVersion, apiClient, regions,
+ *   defaultRegion, modelsFetcher, validateUrl, responsesUrl } — clientId/clientSecret/tokenUrl are
+ *   injected from `oauth` automatically (single source); declare them in `oauth`, not here.
+ *
+ * OAuthConfig: { clientId, authorizeUrl, tokenUrl, deviceCodeUrl, refreshUrl, scope|scopes, redirectUri,
+ *   callbackPath, fixedPort, codeChallengeMethod, extraParams, refresh:{encoding,scope}, refreshLeadMs,
+ *   userInfoUrl }.
+ *
+ * MediaConfig: { serviceKinds:[...], ttsConfig, sttConfig, embeddingConfig, imageConfig,
+ *   searchViaChat:{defaultModel,pricingUrl}, hiddenKinds } — each *Config: {baseUrl,authType,authHeader,
+ *   format,defaultModel,models:[{id,name,dimensions?}]}.
+ */
+
+// Shared transport defaults — provider only overrides fields that differ.
+// NOTE: runtime (index.js buildTransport) only re-applies `format`; the rest documents the contract
+// and feeds the (currently unwired) resolveProvider(). Adding keys here does NOT change PROVIDERS.
+export const PROVIDER_DEFAULTS = {
+  baseUrl: "",
+  format: "openai",
+  headers: {},
+  auth: { header: "Authorization", scheme: "bearer", source: ["accessToken", "apiKey"] },
+  forceStream: false,
+  urlSuffix: "",
+  quirks: {},
+  passthroughModels: false,
+  retry: DEFAULT_RETRY_CONFIG,
+  timeoutMs: FETCH_CONNECT_TIMEOUT_MS,
+  executor: "default"
+};
+
+// Default endpoints per format (provider only overrides what differs)
+export const ENDPOINT_DEFAULTS = {
+  openai: { chat: "/chat/completions", test: "/models", models: "/models" },
+  claude: { chat: "/messages", test: "/models", countTokens: "/messages/count_tokens" },
+  gemini: { chat: "/{model}:streamGenerateContent", models: "/models", test: "/models" }
+};
+
+// Deep-merge a provider entry over PROVIDER_DEFAULTS (defensive for missing transport)
+export function resolveProvider(entry) {
+  const transport = (entry && entry.transport) || {};
+  return {
+    ...PROVIDER_DEFAULTS,
+    ...transport,
+    headers: { ...PROVIDER_DEFAULTS.headers, ...transport.headers },
+    auth: { ...PROVIDER_DEFAULTS.auth, ...transport.auth },
+    quirks: { ...PROVIDER_DEFAULTS.quirks, ...transport.quirks },
+    retry: { ...PROVIDER_DEFAULTS.retry, ...transport.retry }
+  };
+}

package/app/open-sse/providers/shared.js

@@ -0,0 +1,67 @@
+import { platform, arch } from "os";
+
+// === OS/Arch helpers (Stainless fingerprint) ===
+export function mapStainlessOs() {
+  switch (platform()) {
+    case "darwin": return "MacOS";
+    case "win32": return "Windows";
+    case "linux": return "Linux";
+    case "freebsd": return "FreeBSD";
+    default: return `Other::${platform()}`;
+  }
+}
+
+export function mapStainlessArch() {
+  switch (arch()) {
+    case "x64": return "x64";
+    case "arm64": return "arm64";
+    case "ia32": return "x86";
+    default: return `other::${arch()}`;
+  }
+}
+
+// Anthropic API version (single source — reused across claude-format providers/executors)
+export const ANTHROPIC_API_VERSION = "2023-06-01";
+
+// Shared Claude-compatible API headers (reused across claude-format providers)
+export const CLAUDE_API_HEADERS = {
+  "Anthropic-Version": ANTHROPIC_API_VERSION,
+  "Anthropic-Beta": "claude-code-20250219,interleaved-thinking-2025-05-14"
+};
+
+// Full Claude CLI fingerprint — required by providers that gate on client identity (e.g. agentrouter)
+export const CLAUDE_CLI_SPOOF_HEADERS = {
+  "Anthropic-Version": ANTHROPIC_API_VERSION,
+  "Anthropic-Beta": "claude-code-20250219,oauth-2025-04-20,interleaved-thinking-2025-05-14,context-management-2025-06-27,prompt-caching-scope-2026-01-05,advanced-tool-use-2025-11-20,effort-2025-11-24,structured-outputs-2025-12-15,fast-mode-2026-02-01,redact-thinking-2026-02-12,token-efficient-tools-2026-03-28",
+  "Anthropic-Dangerous-Direct-Browser-Access": "true",
+  "User-Agent": "claude-cli/2.1.92 (external, sdk-cli)",
+  "X-App": "cli",
+  "X-Stainless-Helper-Method": "stream",
+  "X-Stainless-Retry-Count": "0",
+  "X-Stainless-Runtime-Version": "v24.14.0",
+  "X-Stainless-Package-Version": "0.80.0",
+  "X-Stainless-Runtime": "node",
+  "X-Stainless-Lang": "js",
+  "X-Stainless-Arch": mapStainlessArch(),
+  "X-Stainless-Os": mapStainlessOs(),
+  "X-Stainless-Timeout": "600"
+};
+
+// Shared baseUrls
+export const KIMI_CODING_BASE_URL = "https://api.kimi.com/coding/v1/messages";
+
+// Default base for dynamic compat providers (openai-compatible-* / anthropic-compatible-*) when user gives no baseUrl
+export const OPENAI_COMPAT_BASE = "https://api.openai.com/v1";
+export const ANTHROPIC_COMPAT_BASE = "https://api.anthropic.com/v1";
+
+// Antigravity OAuth client credentials (public CLI client — duplicated in usage.js + src/lib/oauth)
+export const ANTIGRAVITY_OAUTH_CLIENT = {
+  clientId: "1071006060591-tmhssin2h21lcre235vtolojh4g403ep.apps.googleusercontent.com",
+  clientSecret: "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"
+};
+
+// Gemini (Google) OAuth client credentials (public CLI client — shared by gemini, gemini-cli, src/lib/oauth)
+export const GOOGLE_OAUTH_CLIENT = {
+  clientId: "681255809395-oo8ft2oprdrnp9e3aqf6av3hmdib135j.apps.googleusercontent.com",
+  clientSecret: "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"
+};

package/app/open-sse/rtk/caveman.js

@@ -1,100 +1,9 @@
 // Caveman injector: appends a caveman-style instruction into the system message
 // of the final request body, just before it is dispatched to the provider executor.
-// Dispatches by format so it works for both translated and native-passthrough flows.
 
-import { FORMATS } from "../translator/formats.js";
+import { injectSystemPrompt } from "./systemInject.js";
 import { CAVEMAN_PROMPTS } from "./cavemanPrompts.js";
 
-const SEP = "\n\n";
-
 export function injectCaveman(body, format, level) {
-  const prompt = CAVEMAN_PROMPTS[level];
-  if (!body || !prompt) return;
-
-  switch (format) {
-    case FORMATS.CLAUDE:
-      injectClaudeSystem(body, prompt);
-      return;
-    case FORMATS.GEMINI:
-    case FORMATS.GEMINI_CLI:
-    case FORMATS.VERTEX:
-    case FORMATS.ANTIGRAVITY:
-      // Antigravity wraps Gemini shape in body.request → injectGeminiSystem handles it
-      injectGeminiSystem(body, prompt);
-      return;
-    default:
-      // OpenAI and OpenAI-shaped formats (responses/codex/cursor/kiro/ollama)
-      injectMessagesSystem(body, prompt);
-  }
-}
-
-// OpenAI-shaped: messages[] (chat) or input[] (responses) or instructions (responses string)
-function injectMessagesSystem(body, prompt) {
-  // OpenAI Responses API: top-level string field
-  if (typeof body.instructions === "string") {
-    body.instructions = body.instructions
-      ? `${body.instructions}${SEP}${prompt}`
-      : prompt;
-    return;
-  }
-
-  const arr = Array.isArray(body.messages) ? body.messages
-    : Array.isArray(body.input) ? body.input
-    : null;
-  if (!arr) return;
-
-  const idx = arr.findIndex(m => m && (m.role === "system" || m.role === "developer"));
-  if (idx >= 0) {
-    appendToOpenAIMessage(arr[idx], prompt);
-  } else {
-    arr.unshift({ role: "system", content: prompt });
-  }
-}
-
-function appendToOpenAIMessage(msg, prompt) {
-  if (typeof msg.content === "string") {
-    msg.content = `${msg.content}${SEP}${prompt}`;
-  } else if (Array.isArray(msg.content)) {
-    // Responses-style array of parts {type:"input_text"|"text", text}
-    msg.content.push({ type: "input_text", text: prompt });
-  } else {
-    msg.content = prompt;
-  }
-}
-
-// Claude shape: body.system as string | array of {type:"text", text}
-// Insert before the last cache_control block to keep caveman inside the cached prefix.
-function injectClaudeSystem(body, prompt) {
-  if (typeof body.system === "string" && body.system.length > 0) {
-    body.system = `${body.system}${SEP}${prompt}`;
-    return;
-  }
-  if (Array.isArray(body.system)) {
-    const block = { type: "text", text: prompt };
-    let lastCacheIdx = -1;
-    for (let i = body.system.length - 1; i >= 0; i--) {
-      if (body.system[i]?.cache_control) { lastCacheIdx = i; break; }
-    }
-    if (lastCacheIdx >= 0) {
-      body.system.splice(lastCacheIdx, 0, block);
-    } else {
-      body.system.push(block);
-    }
-    return;
-  }
-  body.system = prompt;
-}
-
-// Gemini shape: body.system_instruction | body.systemInstruction | body.request.systemInstruction
-// Each shape: { parts: [{ text }] }
-function injectGeminiSystem(body, prompt) {
-  const target = body.request && typeof body.request === "object" ? body.request : body;
-  const useSnake = Object.prototype.hasOwnProperty.call(target, "system_instruction");
-  const key = useSnake ? "system_instruction" : "systemInstruction";
-  const sys = target[key];
-  if (sys && Array.isArray(sys.parts)) {
-    sys.parts.push({ text: prompt });
-    return;
-  }
-  target[key] = { parts: [{ text: prompt }] };
+  injectSystemPrompt(body, format, CAVEMAN_PROMPTS[level]);
 }