@gajae-code/ai 0.11.11 → 0.12.1

package/CHANGELOG.md

@@ -2,12 +2,54 @@
 
 ## [Unreleased]
 
+## [0.12.1] - 2026-07-29
+
+### Fixed
+
+- Lazy-stream first-event timeouts now abort with `FirstEventTimeoutError` so `transportFailure.providerCode` is `stream_first_event_timeout` on the outer watchdog path shared by all bundled providers via `createLazyStream`. Idle stalls remain bare `Error`s (distinct class intentionally) (#3496).
+
+- Provider streams now surface first-event watchdog expiry as a typed timeout so callers can apply bounded retry policy without parsing error prose.
+- Codex websocket first-event timeouts now discard the timed-out connection before the outer retry/fallback layer handles the typed failure, preventing late frames from the abandoned request from being consumed by the replayed turn.
+- Codex named-tool requests now recognize provider `Tool choice '<name>' not found in 'tools' parameter` errors as runtime capability failures and retry once without forcing the choice.
+- The Kimi OAuth host (`KIMI_CODE_OAUTH_HOST` / `KIMI_OAUTH_HOST`) is now resolved from trusted environment sources only. That host receives the device-authorization request, the authorization-code exchange, and the refresh call that carries the existing refresh token, so reading it through the merged view that includes the caller's `cwd/.env` let a repository redirect the login flow and collect the user's Kimi credentials. Resolution now uses the non-project resolver; shell and user-level configuration is unchanged.
+- The documented `GJC_NO_STRICT` environment variable now takes effect. `adaptSchemaForStrict` read only the legacy `PI_NO_STRICT`, so an operator hitting a provider that rejects strict function schemas set the documented name and strict mode stayed on. Both names are honoured, canonical name first, and `GJC_NO_STRICT` is now listed in the environment-variable reference rather than only in the schema-normalisation note.
+- The documented `GJC_AUTH_NO_BORROW` environment variable now takes effect. Only the legacy `PI_AUTH_NO_BORROW` was read, so an operator who followed the documentation to disable macOS native-app token borrowing still had a JWT read out of the Perplexity desktop application during login. Both names are now honoured, and the contract stays presence-based as documented so that setting it to `0` cannot silently re-enable borrowing.
+- The Azure client's `AZURE_OPENAI_API_KEY` fallback is now resolved from trusted environment sources only. It read the merged view that includes the caller's `cwd/.env`, so a repository could supply the credential the client authenticates with; provider credential resolution is documented as excluding the project `.env`, and this fallback now matches. An explicit caller-supplied key still takes precedence, and shell / user-level configuration is unchanged.
+- Anthropic and Ollama tool calls cut off by an output-token limit are now marked incomplete before dispatch, so repaired partial JSON is rejected instead of executing with truncated arguments.
+- The Anthropic "thinking blocks in the latest assistant message cannot be modified" 400 now escalates its one-shot replay repair. The error names the latest assistant message but its cited `messages.N.content.M` path can point at an earlier replayed turn, so the latest-only repair was rejected identically and killed the turn; recovery now retries once more with thinking dropped from every replayed assistant message.
+- Anthropic adaptive-thinking `display` support is now decided by the canonical model-version parser instead of a provider-local `claude-opus-(\d+)-(\d+)` regex. The regex only matched two-component ids, so a single-component alias such as `claude-opus-5` was classified as pre-4.7 while its dated snapshot `claude-opus-5-20260101` was not: the alias sent `thinking: { type: "adaptive" }` without `display: "summarized"`, additionally requested the `interleaved-thinking-2025-05-14` beta, and had its returned thinking blocks recorded as raw rather than summarized. Both Anthropic and Bedrock providers now share `supportsAnthropicAdaptiveThinkingDisplay`, so alias and dated ids of the same model send an identical request shape.
+- Anthropic requests that force a tool choice no longer replay signed thinking blocks. Forcing `tool_choice` strips `thinking` from the request (the API rejects the combination), but the converted history still carried native `thinking`/`redacted_thinking` blocks from thinking-enabled turns, so eager tool-forcing turns (e.g. the todo bootstrap) sent a request whose history contradicted its own thinking setting and drew a 400. The replay now degrades in the same rebuild; the forced request trades its prompt-cache prefix for a shape the API accepts.
+
+## [0.12.0] - 2026-07-28
+
+### Added
+
+- Added first-class support for **BizRouter**, an OpenAI-compatible Korean enterprise LLM gateway. Registers the `bizrouter` provider descriptor, `/login` entry (API-key paste validated against `https://api.bizrouter.ai/v1/models`), `BIZROUTER_API_KEY` environment resolution, and bundled `models.json` seed models. Models are discovered dynamically from `GET /v1/models` (base URL `https://api.bizrouter.ai/v1`).
+### Fixed
+
+- Anthropic subscription OAuth requests now use the current Claude Code compatibility attribution (`2.1.219`, `sdk-cli`) instead of the stale `2.1.63` CLI fingerprint that Anthropic can misclassify as extra usage.
+- Connection failures now name the transport code and the target URL. Bun reports DNS and socket failures as a bare `Error` whose message is a standalone hint ("Was there a typo in the url or port?", "Unable to connect. Is the computer able to access the url?") and keeps the actionable facts on `code` and `path`, but only `message` reached the assistant message. A provider outage, a local DNS failure, and a mistyped custom base URL therefore all rendered as the same context-free sentence with no host in it. Such failures now read `... (transport=FailedToOpenSocket url=https://chatgpt.com/backend-api/codex/responses)`; the URL is reduced to origin and path so a key carried in the query string is not surfaced.
+
+### Documentation
+
+- `docs/environment-variables.md` now names the Anthropic Foundry gateway variables that are actually read: `CLAUDE_CODE_USE_FOUNDRY`, `CLAUDE_CODE_CLIENT_CERT`, and `CLAUDE_CODE_CLIENT_KEY`. The page advertised `ANTHROPIC_MODEL_CODE_*` spellings that no code path reads, so an operator following it could not enable Foundry mode at all, and the mTLS client material was silently ignored.
+
 ## [0.11.11] - 2026-07-26
 
 ### Fixed
 
+- The Kimi usage endpoint base (`KIMI_CODE_BASE_URL`) is now resolved from trusted environment sources only. That base becomes the URL the usage request sends `Authorization: Bearer <accessToken>` to, so reading it through the merged view that includes the caller's `cwd/.env` let a repository collect the user's Kimi access token. An explicit caller-supplied base URL still takes precedence, and shell / user-level configuration is unchanged.
+- The Gemini CLI compatibility version used in the outbound `User-Agent` is refreshed from `0.50.0` to `0.52.0`, matching the current upstream release. The repository ships `check-spoofed-versions` for exactly this, but that check is not wired into CI, so the value had drifted two minor releases behind.
+- The OpenAI and Azure endpoint decisions are now resolved from trusted environment sources only: `OPENAI_BASE_URL` (streaming responses, completions, and the model manager) and `AZURE_OPENAI_BASE_URL`. `Bun.env` is `process.env` and the env module merges the caller's `cwd/.env` into it, so a repository could previously plant a `.env` that redirected authenticated requests; the two provider paths already reached for `$inheritedEnv` but re-admitted the project `.env` through a fallback. Resolution now goes through the non-project resolver (launching shell plus GJC/user-owned `.env` files); shell and user-level configuration is unchanged.
+- The OpenAI and Azure endpoint decisions are now resolved from trusted environment sources only: `OPENAI_BASE_URL` (streaming responses, completions, and the model manager), `AZURE_OPENAI_BASE_URL`, and `AZURE_OPENAI_RESOURCE_NAME` (the alternate constructor for the same Azure host). `Bun.env` is `process.env` and the env module merges the caller's `cwd/.env` into it, so a repository could previously plant a `.env` that redirected authenticated requests; the two provider paths already reached for `$inheritedEnv` but re-admitted the project `.env` through a fallback. Resolution now goes through the non-project resolver (launching shell plus GJC/user-owned `.env` files); shell and user-level configuration is unchanged.
+- Google credential material is now resolved from trusted environment sources only: the `GOOGLE_APPLICATION_CREDENTIALS` service-account / authorized-user file path used by the ADC loader, and `GOOGLE_CLOUD_API_KEY` used as the Vertex API key. Both were read through the merged view that includes the caller's `cwd/.env`, so a repository could ship a key file and point the agent at it, making it authenticate to Google as an identity the repository chose. `stream.ts` already resolved the same ADC variable through the non-project resolver; the two now agree. An explicit caller-supplied API key still takes precedence.
+- The Grok usage token fallback (`GROK_CLI_OAUTH_TOKEN`) is now resolved from trusted environment sources only. It authenticates the billing/usage call, and reading it through the merged view that includes the caller's `cwd/.env` let a repository decide which account that call ran against. Stored credentials keep precedence, and shell / user-level configuration is unchanged.
+- The Vertex AI location (`GOOGLE_CLOUD_LOCATION`) can no longer redirect authenticated requests off Google. It is interpolated into the request host (`${location}-aiplatform.googleapis.com`), so a value containing `/` terminated the authority component and sent the Google access token to an arbitrary origin — and it was read through the merged view that includes the caller's `cwd/.env`. It now resolves from trusted sources only and must be a region label; `GOOGLE_CLOUD_PROJECT` / `GCLOUD_PROJECT` moved to the same trusted resolver.
+- HTTP 400 request dumps are now bounded. Every 400 wrote a file containing the full sanitized request body and nothing ever removed one, so the directory grew without limit — a developer machine reached 27,249 files totalling 7.0 GB, which was 96% of everything under `~/.gjc`. The newest 50 are retained, matching the bounded retention the rotating application log already uses, and pruning stays best-effort so diagnostics never turn a request failure into a second failure.
 - Anthropic `ping` keepalives no longer reset stream progress, so responses that stop producing content now reach the idle timeout instead of hanging indefinitely.
+- The Anthropic endpoint decision is now resolved from trusted environment sources only: `ANTHROPIC_BASE_URL`, `FOUNDRY_BASE_URL`, `ZCODE_PLAN_ANTHROPIC_BASE_URL`, and the `CLAUDE_CODE_USE_FOUNDRY` mode switch. `Bun.env` is `process.env` and the env module merges the caller's `cwd/.env` into it, so a repository could previously plant a `.env` that redirected authenticated Anthropic requests — the resolved base URL becomes `${baseUrl}/v1/messages` while the headers carry the API key or OAuth token. Resolution now goes through the non-project resolver (launching shell plus GJC/user-owned `.env` files); shell and user-level configuration is unchanged.
 - The documented `GJC_OPENAI_STREAM_IDLE_TIMEOUT_MS` environment variable now takes effect: the stream-watchdog idle-timeout helpers resolve it GJC-first before the legacy `PI_OPENAI_STREAM_IDLE_TIMEOUT_MS` / `PI_STREAM_IDLE_TIMEOUT_MS` aliases (previously only the `PI_`-prefixed names were read, so setting the documented GJC name was a silent no-op).
+- The documented OpenAI-code provider knobs now take effect: `GJC_OPENAI_CODE_DEBUG`, `GJC_OPENAI_CODE_WEBSOCKET`, `GJC_OPENAI_CODE_WEBSOCKET_IDLE_TIMEOUT_MS`, `GJC_OPENAI_CODE_WEBSOCKET_RETRY_BUDGET`, and `GJC_OPENAI_CODE_WEBSOCKET_RETRY_DELAY_MS` are resolved GJC-first ahead of the legacy `PI_CODEX_*` names. The Codex → OpenAI-code rename had updated the documentation but not the reads, so every documented name was a silent no-op.
 
 ## [0.11.9] - 2026-07-24
 ### Fixed

package/README.md

@@ -70,6 +70,7 @@ Unified LLM API with automatic model discovery, provider configuration, token an
 - **Xiaomi MiMo** (requires `XIAOMI_API_KEY`)
 - **ZenMux** (requires `ZENMUX_API_KEY`)
 - **OpenGateway by Sionic AI** (requires `OPENGATEWAY_API_KEY`)
+- **BizRouter** (requires `BIZROUTER_API_KEY`)
 - **Qwen Portal** (supports `QWEN_OAUTH_TOKEN` or `QWEN_PORTAL_API_KEY`)
 - **Cloudflare AI Gateway** (requires `CLOUDFLARE_AI_GATEWAY_API_KEY` and provider-specific gateway base URL)
 - **Ollama** (local OpenAI-compatible runtime; optional `OLLAMA_API_KEY`)
@@ -956,6 +957,7 @@ In Node.js environments, you can set environment variables to avoid passing API
 | Xiaomi MiMo    | `XIAOMI_API_KEY`                                                             |
 | ZenMux         | `ZENMUX_API_KEY`                                                             |
 | OpenGateway    | `OPENGATEWAY_API_KEY`                                                        |
+| BizRouter      | `BIZROUTER_API_KEY`                                                          |
 | vLLM           | `VLLM_API_KEY`                                                               |
 | Cloudflare AI Gateway | `CLOUDFLARE_AI_GATEWAY_API_KEY`                                      |
 | GitHub Copilot | `COPILOT_GITHUB_TOKEN` or `GH_TOKEN` or `GITHUB_TOKEN`                      |
@@ -980,6 +982,7 @@ Provider endpoint defaults for the current OpenAI-compatible integrations:
 - ZenMux (OpenAI): `https://zenmux.ai/api/v1`
 - ZenMux (Anthropic models): `https://zenmux.ai/api/anthropic`
 - OpenGateway by Sionic AI: `https://apis.opengateway.ai/v1`
+- BizRouter: `https://api.bizrouter.ai/v1`
 - vLLM: `http://127.0.0.1:8000/v1`
 - Ollama: local OpenAI-compatible runtime (`http://127.0.0.1:11434/v1`)
 - Ollama Cloud: native Ollama API host (`https://ollama.com/api`, configured here as base URL `https://ollama.com`)

package/dist/types/model-thinking.d.ts

@@ -72,3 +72,18 @@ export declare function mapEffortToAnthropicAdaptiveEffort<TApi extends Api>(mod
  * - Thinking content is omitted by default (needs display: "summarized")
  */
 export declare function hasOpus47ApiRestrictions(modelId: string): boolean;
+/**
+ * Adaptive thinking `display` is supported starting with Anthropic Opus 4.7.
+ * Older adaptive-thinking models (Opus 4.6, Sonnet 4.6+) reject the field.
+ * Fable (5+) postdates Opus 4.7, accepts `display`, and defaults it to
+ * "omitted" — thinking tokens are billed but no content streams back — so it
+ * must opt in like Opus 4.7+ (issue #2791).
+ *
+ * Shares `hasOpus47ApiRestrictions` version parsing on purpose: the two
+ * predicates describe the same API generation, and a private `claude-opus-(\d+)-(\d+)`
+ * regex silently disagreed with it for single-component aliases (`claude-opus-5`
+ * matched nothing while `claude-opus-5-20260101` matched), so the same model sent
+ * a different thinking shape and beta set depending on which id string was used.
+ * Bedrock region/inference-profile prefixes are handled by the canonical parser.
+ */
+export declare function supportsAnthropicAdaptiveThinkingDisplay(modelId: string): boolean;

package/dist/types/provider-models/openai-compat.d.ts

@@ -27,6 +27,8 @@ export interface OpenAIModelManagerConfig {
     apiKey?: string;
     baseUrl?: string;
 }
+/** Test seam: the model-manager base URL as resolved from trusted env. */
+export declare function resolveOpenAIModelManagerBaseUrlForTest(config?: OpenAIModelManagerConfig): string;
 export declare function openaiModelManagerOptions(config?: OpenAIModelManagerConfig): ModelManagerOptions<"openai-responses">;
 export interface GroqModelManagerConfig {
     apiKey?: string;
@@ -121,6 +123,11 @@ export interface OpenGatewayModelManagerConfig {
  * the OpenAI-compatible `/v1/models` endpoint.
  */
 export declare function opengatewayModelManagerOptions(config?: OpenGatewayModelManagerConfig): ModelManagerOptions<"openai-completions">;
+export interface BizRouterModelManagerConfig {
+    apiKey?: string;
+    baseUrl?: string;
+}
+export declare function bizrouterModelManagerOptions(config?: BizRouterModelManagerConfig): ModelManagerOptions<"openai-completions">;
 export interface KiloModelManagerConfig {
     apiKey?: string;
     baseUrl?: string;

package/dist/types/providers/anthropic.d.ts

@@ -50,7 +50,8 @@ export declare function isAnthropicThinkingBlockMutationError(error: unknown): b
  * than only the latest one.
  */
 export declare function isAnthropicThinkingSignatureInvalidError(error: unknown): boolean;
-export declare const claudeCodeVersion = "2.1.63";
+export declare const claudeCodeVersion = "2.1.219";
+export declare const claudeCodeEntrypoint = "sdk-cli";
 export declare const claudeToolPrefix: string;
 export declare const claudeCodeSystemInstruction = "You are a Claude agent, built on Anthropic's Claude Agent SDK.";
 export declare function mapStainlessOs(platform: string): "MacOS" | "Windows" | "Linux" | "FreeBSD" | `Other::${string}`;

package/dist/types/providers/azure-openai-responses.d.ts

@@ -1,4 +1,4 @@
-import type { ServiceTier, StreamFunction, StreamOptions, ToolChoice } from "../types";
+import type { Model, ServiceTier, StreamFunction, StreamOptions, ToolChoice } from "../types";
 export interface AzureOpenAIResponsesOptions extends StreamOptions {
     reasoning?: "minimal" | "low" | "medium" | "high" | "xhigh" | "max";
     reasoningSummary?: "auto" | "detailed" | "concise" | null;
@@ -13,3 +13,10 @@ export interface AzureOpenAIResponsesOptions extends StreamOptions {
  * Generate function for Azure OpenAI Responses API
  */
 export declare const streamAzureOpenAIResponses: StreamFunction<"azure-openai-responses">;
+/** Test seam: the Azure endpoint config as resolved from trusted env. */
+export declare function resolveAzureConfigForTest(model: Model<"azure-openai-responses">, options?: AzureOpenAIResponsesOptions): {
+    baseUrl: string;
+    apiVersion: string;
+};
+/** Test seam: the client API key as resolved from a caller value plus trusted env. */
+export declare function resolveAzureClientApiKeyForTest(apiKey: string): string | undefined;

package/dist/types/providers/google-auth.d.ts

@@ -12,6 +12,8 @@
  * (default 60s). Concurrent callers waiting on a refresh share the same in-flight promise.
  */
 import type { FetchImpl } from "../types";
+/** Test seam: the ADC credentials file path as resolved from trusted env. */
+export declare function resolveAdcCredentialsPathForTest(): string | undefined;
 /**
  * Returns a Bearer access token suitable for the `Authorization` header on Vertex AI calls.
  * The token is cached in module scope and refreshed `GOOGLE_VERTEX_REFRESH_SKEW_MS` ms before it expires.

package/dist/types/providers/google-gemini-headers.d.ts

@@ -5,7 +5,7 @@
  */
 export declare const GEMINI_CLI_VERSION_ENV = "GJC_AI_GEMINI_CLI_VERSION";
 export declare const LEGACY_GEMINI_CLI_VERSION_ENV = "PI_AI_GEMINI_CLI_VERSION";
-export declare const DEFAULT_GEMINI_CLI_VERSION = "0.50.0";
+export declare const DEFAULT_GEMINI_CLI_VERSION = "0.52.0";
 export declare function getGeminiCliUserAgent(modelId?: string): string;
 export declare const getGeminiCliHeaders: (modelId?: string) => {
     "User-Agent": string;

package/dist/types/providers/google-vertex.d.ts

@@ -5,3 +5,7 @@ export interface GoogleVertexOptions extends GoogleSharedStreamOptions {
     location?: string;
 }
 export declare const streamGoogleVertex: StreamFunction<"google-vertex">;
+/** Test seam: the Vertex API key as resolved from options plus trusted env. */
+export declare function resolveVertexApiKeyForTest(options?: GoogleVertexOptions): string | undefined;
+/** Test seam: the Vertex location as resolved from options plus trusted env. */
+export declare function resolveVertexLocationForTest(options?: GoogleVertexOptions): string;

package/dist/types/providers/openai-codex-responses.d.ts

@@ -10,6 +10,10 @@ export interface OpenAICodexResponsesOptions extends StreamOptions {
     preferWebsockets?: boolean;
     serviceTier?: ServiceTier;
 }
+/** Maps a canonical tool name to the name Codex accepts on the wire. */
+export declare function codexToolWireName(name: string): string;
+/** Maps a Codex wire tool name back to the canonical harness tool name. */
+export declare function codexToolCanonicalName(wireName: string): string;
 type CodexTransport = "sse" | "websocket";
 export interface OpenAICodexWebSocketDebugStats {
     fullContextRequests: number;

package/dist/types/providers/openai-completions.d.ts

@@ -1,6 +1,8 @@
 import type { ChatCompletionMessageParam } from "openai/resources/chat/completions";
 import { type AssistantMessage, type Context, type Model, type ServiceTier, type StreamFunction, type StreamOptions, type ToolChoice } from "../types";
 import { type ResolvedOpenAICompat } from "./openai-completions-compat";
+/** Test seam: the provider base URL as resolved from trusted env. */
+export declare function resolveOpenAICompletionsBaseUrlForTest(baseUrl: string | undefined, authCredentialType: "api_key" | "oauth" | undefined): string;
 /**
  * Identify "real progress" stream chunks vs. keepalives, role-only preambles,
  * and empty `{choices:[]}` no-ops emitted by some OpenAI-compatible endpoints.

package/dist/types/providers/openai-responses.d.ts

@@ -13,6 +13,8 @@ export interface OpenAIResponsesOptions extends StreamOptions {
      */
     strictResponsesPairing?: boolean;
 }
+/** Test seam: the provider base URL as resolved from trusted env. */
+export declare function resolveOpenAIProviderBaseUrlForTest(baseUrl: string | undefined, authCredentialType: "api_key" | "oauth" | undefined): string;
 /**
  * Generate function for OpenAI Responses API
  */

package/dist/types/types.d.ts

@@ -51,7 +51,7 @@ export interface ThinkingConfig {
     /** Provider-specific transport used to encode the selected effort. */
     mode: ThinkingControlMode;
 }
-export type KnownProvider = "alibaba-token-plan" | "amazon-bedrock" | "azure-openai" | "anthropic" | "google" | "google-gemini-cli" | "google-antigravity" | "google-vertex" | "openai" | "openai-codex" | "kimi-code" | "minimax-code" | "minimax-code-cn" | "github-copilot" | "fireworks" | "firepass" | "fugu" | "gitlab-duo" | "cursor" | "deepseek" | "deepinfra" | "xai" | "groq" | "cerebras" | "openrouter" | "kilo" | "vercel-ai-gateway" | "zai" | "glm-zcode" | "mistral" | "minimax" | "opencode-go" | "opencode-zen" | "opengateway" | "synthetic" | "cloudflare-ai-gateway" | "huggingface" | "litellm" | "moonshot" | "nvidia" | "nanogpt" | "ollama" | "ollama-cloud" | "qianfan" | "qwen-portal" | "together" | "venice" | "vllm" | "xiaomi" | "xiaomi-token-plan-sgp" | "xiaomi-token-plan-ams" | "xiaomi-token-plan-cn" | "zenmux" | "lm-studio";
+export type KnownProvider = "alibaba-token-plan" | "amazon-bedrock" | "azure-openai" | "anthropic" | "google" | "google-gemini-cli" | "google-antigravity" | "google-vertex" | "openai" | "openai-codex" | "kimi-code" | "minimax-code" | "minimax-code-cn" | "github-copilot" | "fireworks" | "firepass" | "fugu" | "gitlab-duo" | "cursor" | "deepseek" | "deepinfra" | "xai" | "groq" | "cerebras" | "openrouter" | "kilo" | "vercel-ai-gateway" | "zai" | "glm-zcode" | "mistral" | "minimax" | "opencode-go" | "opencode-zen" | "opengateway" | "bizrouter" | "synthetic" | "cloudflare-ai-gateway" | "huggingface" | "litellm" | "moonshot" | "nvidia" | "nanogpt" | "ollama" | "ollama-cloud" | "qianfan" | "qwen-portal" | "together" | "venice" | "vllm" | "xiaomi" | "xiaomi-token-plan-sgp" | "xiaomi-token-plan-ams" | "xiaomi-token-plan-cn" | "zenmux" | "lm-studio";
 export type Provider = KnownProvider | string;
 import type { Effort } from "./model-thinking";
 /** Token budgets for each thinking level (token-based providers only) */

package/dist/types/usage/grok-cli.d.ts

@@ -1,10 +1,12 @@
-import type { CredentialRankingStrategy, UsageProvider } from "../usage";
+import type { CredentialRankingStrategy, UsageFetchParams, UsageProvider } from "../usage";
 interface BillingUsage {
     monthlyLimit: number;
     used: number;
     billingPeriodEnd: string;
 }
 export declare function parseGrokCliBillingUsage(payload: unknown): BillingUsage;
+/** Test seam: the usage access token as resolved from a credential plus trusted env. */
+export declare function resolveGrokAccessTokenForTest(params: UsageFetchParams): string | undefined;
 export declare const grokCliUsageProvider: UsageProvider;
 export declare const grokCliRankingStrategy: CredentialRankingStrategy;
 export {};

package/dist/types/usage/kimi.d.ts

@@ -1,2 +1,4 @@
 import type { UsageProvider } from "../usage";
+/** Test seam: the usage base URL as resolved from a caller value plus trusted env. */
+export declare function normalizeKimiUsageBaseUrlForTest(baseUrl?: string): string;
 export declare const kimiUsageProvider: UsageProvider;

package/dist/types/utils/anthropic-auth.d.ts

@@ -4,6 +4,14 @@ export interface AnthropicAuthConfig {
     baseUrl: string;
     isOAuth: boolean;
 }
+/**
+ * Resolve the Anthropic base URL from the environment.
+ *
+ * Trusted sources only: the result becomes the request URL that carries the
+ * Anthropic API key / OAuth token, so whatever can set it can redirect
+ * authenticated traffic. `$env` merges the caller's `cwd/.env`, so reading it
+ * there would let repository content choose where credentials are sent.
+ */
 export declare function resolveAnthropicBaseUrlFromEnv(): string | undefined;
 /**
  * Checks if a token is an OAuth token by looking for sk-ant-oat prefix.

package/dist/types/utils/fallback-transport.d.ts

@@ -3,6 +3,8 @@ export interface FallbackTrigger {
     class: FallbackTriggerClass;
     retryAfterMs?: number;
 }
+/** Stable code for streams that time out before producing semantic progress. */
+export declare const STREAM_FIRST_EVENT_TIMEOUT_PROVIDER_CODE = "stream_first_event_timeout";
 export type TransportHeaders = Headers | Record<string, string | undefined>;
 /**
  * Structured facts from an upstream HTTP or transport failure. Retry decisions

package/dist/types/utils/foundry.d.ts

@@ -1 +1,11 @@
+/**
+ * Whether Anthropic requests run in Foundry gateway mode.
+ *
+ * Resolved from trusted environment sources only. Enabling Foundry switches the
+ * request base URL and injects TLS client material, so whatever can set this
+ * redirects authenticated traffic. `$env` merges the caller's `cwd/.env`, so
+ * reading it there would let repository content flip the mode; resolve it the
+ * same way the credentials themselves are (launching shell plus GJC/user-owned
+ * `.env` files, never the project `.env`).
+ */
 export declare function isFoundryEnabled(): boolean;

package/dist/types/utils/http-inspector.d.ts

@@ -17,7 +17,30 @@ export type CapturedHttpErrorResponse = {
 export declare function isModelUnavailableError(message: string, error: unknown): boolean;
 /** Actionable guidance for selecting an available model/provider. */
 export declare function formatModelUnavailableGuidance(dump: RawHttpRequestDump | undefined): string;
+/** Directory holding the retained HTTP 400 dumps. */
+export declare function httpRequestDumpDir(): string;
+/**
+ * Drop the oldest dumps beyond the cap. Best-effort: diagnostics must never turn
+ * a request failure into a second failure, so every step swallows its error.
+ *
+ * File names are `${Date.now()}-${hash}.json`, so a lexical sort is chronological
+ * for the millisecond timestamps this writer produces.
+ */
+export declare function pruneHttpRequestDumps(dir?: string): Promise<number>;
 export declare function appendRawHttpRequestDumpFor400(message: string, error: unknown, dump: RawHttpRequestDump | undefined): Promise<string>;
+/**
+ * Name the failed connection when the request never produced an HTTP status.
+ *
+ * Bun raises DNS and socket failures as a bare `Error` whose message is a
+ * standalone hint ("Was there a typo in the url or port?", "Unable to connect.
+ * Is the computer able to access the url?") while the actionable facts live on
+ * `code` and `path`. Those properties are dropped when only `message` reaches
+ * the assistant message, so a provider outage, a local DNS failure, and a
+ * mistyped custom base URL all render as the same context-free sentence.
+ * Appending the code and the target URL tells the user which host failed and
+ * whether the fault is theirs.
+ */
+export declare function appendTransportFailureContext(message: string, error: unknown, rawRequestDump: RawHttpRequestDump | undefined): string;
 export declare function finalizeErrorMessage(error: unknown, rawRequestDump: RawHttpRequestDump | undefined, capturedErrorResponse?: CapturedHttpErrorResponse): Promise<string>;
 export declare function withHttpStatus(error: unknown, status: number): Error;
 /**

package/dist/types/utils/idle-iterator.d.ts

@@ -30,6 +30,10 @@ export declare function getOpenAIStreamIdleTimeoutMs(): number | undefined;
  */
 export declare function getStreamFirstEventTimeoutMs(idleTimeoutMs?: number, fallbackMs?: number): number | undefined;
 export type Watchdog = NodeJS.Timeout | undefined;
+export declare class FirstEventTimeoutError extends Error {
+    readonly providerCode = "stream_first_event_timeout";
+    constructor(message: string);
+}
 /**
  * Starts a watchdog that aborts a request if no first stream event arrives in time.
  * Call `markFirstEventReceived()` as soon as the first event is observed.

package/dist/types/utils/oauth/bizrouter.d.ts

@@ -0,0 +1 @@
+export declare const loginBizRouter: (options: import("./types").OAuthController) => Promise<string>;

package/dist/types/utils/oauth/kimi.d.ts

@@ -2,6 +2,8 @@
  * Kimi Code OAuth flow (device authorization grant)
  */
 import type { OAuthController, OAuthCredentials } from "./types";
+/** Test seam: the OAuth host as resolved from trusted env. */
+export declare function resolveKimiOAuthHostForTest(): string;
 export declare let getKimiCommonHeaders: () => Readonly<{
     "User-Agent": `KimiCLI/${string}`;
     "X-Msh-Platform": "kimi_cli";

package/dist/types/utils/oauth/perplexity.d.ts

@@ -1,9 +1,4 @@
 import type { OAuthController, OAuthCredentials } from "./types";
-/**
- * Login to Perplexity.
- *
- * Tries auto-extraction from the desktop app, then runs HTTP email OTP login.
- *
- * No browser/manual token paste fallback is used.
- */
+/** Test seam: the resolved native-app borrowing opt-out. */
+export declare function authBorrowDisabledForTest(): boolean;
 export declare function loginPerplexity(ctrl: OAuthController): Promise<OAuthCredentials>;

package/dist/types/utils/oauth/types.d.ts

@@ -7,7 +7,7 @@ export type OAuthCredentials = {
     email?: string;
     accountId?: string;
 };
-export type OAuthProvider = "alibaba-token-plan" | "anthropic" | "cerebras" | "cloudflare-ai-gateway" | "cursor" | "deepseek" | "deepinfra" | "fireworks" | "firepass" | "fugu" | "github-copilot" | "google-gemini-cli" | "google-antigravity" | "gitlab-duo" | "huggingface" | "kimi-code" | "kilo" | "kagi" | "litellm" | "lm-studio" | "minimax-code" | "minimax-code-cn" | "moonshot" | "nvidia" | "nanogpt" | "ollama" | "ollama-cloud" | "openai-codex" | "openai-codex-device" | "opencode-go" | "opencode-zen" | "opengateway" | "parallel" | "perplexity" | "qianfan" | "qwen-portal" | "synthetic" | "tavily" | "together" | "venice" | "vercel-ai-gateway" | "vllm" | "xai" | "glm-zcode" | "xiaomi" | "xiaomi-token-plan-sgp" | "xiaomi-token-plan-ams" | "xiaomi-token-plan-cn" | "zenmux" | "zai";
+export type OAuthProvider = "alibaba-token-plan" | "anthropic" | "bizrouter" | "cerebras" | "cloudflare-ai-gateway" | "cursor" | "deepseek" | "deepinfra" | "fireworks" | "firepass" | "fugu" | "github-copilot" | "google-gemini-cli" | "google-antigravity" | "gitlab-duo" | "huggingface" | "kimi-code" | "kilo" | "kagi" | "litellm" | "lm-studio" | "minimax-code" | "minimax-code-cn" | "moonshot" | "nvidia" | "nanogpt" | "ollama" | "ollama-cloud" | "openai-codex" | "openai-codex-device" | "opencode-go" | "opencode-zen" | "opengateway" | "parallel" | "perplexity" | "qianfan" | "qwen-portal" | "synthetic" | "tavily" | "together" | "venice" | "vercel-ai-gateway" | "vllm" | "xai" | "glm-zcode" | "xiaomi" | "xiaomi-token-plan-sgp" | "xiaomi-token-plan-ams" | "xiaomi-token-plan-cn" | "zenmux" | "zai";
 export type OAuthProviderId = OAuthProvider | (string & {});
 export type OAuthPrompt = {
     message: string;

package/package.json

@@ -1,7 +1,7 @@
 {
 	"type": "module",
 	"name": "@gajae-code/ai",
-	"version": "0.11.11",
+	"version": "0.12.1",
 	"description": "Unified LLM API with automatic model discovery and provider configuration",
 	"homepage": "https://gajae-code.com",
 	"author": "Yeachan-Heo and Gajae Code Contributors",
@@ -40,7 +40,7 @@
 	"dependencies": {
 		"@anthropic-ai/sdk": "^0.94.0",
 		"@bufbuild/protobuf": "^2.12.0",
-		"@gajae-code/utils": "0.11.11",
+		"@gajae-code/utils": "0.12.1",
 		"openai": "^6.36.0",
 		"partial-json": "^0.1.7",
 		"zod": "4.4.3"

package/src/auth-storage.ts

@@ -1998,6 +1998,12 @@ export class AuthStorage {
 				await saveApiKeyCredential(apiKey);
 				return;
 			}
+			case "bizrouter": {
+				const { loginBizRouter } = await import("./utils/oauth/bizrouter");
+				const apiKey = await loginBizRouter(ctrl);
+				await saveApiKeyCredential(apiKey);
+				return;
+			}
 			case "opengateway": {
 				const { loginOpenGateway } = await import("./utils/oauth/opengateway");
 				const apiKey = await loginOpenGateway(ctrl);

package/src/cli.ts

@@ -119,6 +119,7 @@ Providers:
   cursor            Cursor (Anthropic, GPT, etc.)
   zenmux            ZenMux
   opengateway       OpenGateway by Sionic AI
+  bizrouter         BizRouter
   ollama-cloud      Ollama Cloud
 
 Examples:

package/src/model-thinking.ts

@@ -350,6 +350,25 @@ export function hasOpus47ApiRestrictions(modelId: string): boolean {
 	return semverGte(parsed.version, "4.7") && parsed.kind === "opus";
 }
 
+/**
+ * Adaptive thinking `display` is supported starting with Anthropic Opus 4.7.
+ * Older adaptive-thinking models (Opus 4.6, Sonnet 4.6+) reject the field.
+ * Fable (5+) postdates Opus 4.7, accepts `display`, and defaults it to
+ * "omitted" — thinking tokens are billed but no content streams back — so it
+ * must opt in like Opus 4.7+ (issue #2791).
+ *
+ * Shares `hasOpus47ApiRestrictions` version parsing on purpose: the two
+ * predicates describe the same API generation, and a private `claude-opus-(\d+)-(\d+)`
+ * regex silently disagreed with it for single-component aliases (`claude-opus-5`
+ * matched nothing while `claude-opus-5-20260101` matched), so the same model sent
+ * a different thinking shape and beta set depending on which id string was used.
+ * Bedrock region/inference-profile prefixes are handled by the canonical parser.
+ */
+export function supportsAnthropicAdaptiveThinkingDisplay(modelId: string): boolean {
+	if (/claude-fable-\d/.test(modelId)) return true;
+	return hasOpus47ApiRestrictions(modelId);
+}
+
 function anthropicModelHasRealXHighEffort<TApi extends Api>(model: ApiModel<TApi>): boolean {
 	if (model.api !== "anthropic-messages") return false;
 	const parsedModel = parseKnownModel(model.id);

package/src/models.json

@@ -4030,6 +4030,78 @@
 			}
 		}
 	},
+	"bizrouter": {
+		"anthropic/claude-sonnet-4.5": {
+			"id": "anthropic/claude-sonnet-4.5",
+			"name": "Anthropic Sonnet 4.5",
+			"api": "openai-completions",
+			"provider": "bizrouter",
+			"baseUrl": "https://api.bizrouter.ai/v1",
+			"reasoning": true,
+			"input": [
+				"text",
+				"image"
+			],
+			"cost": {
+				"input": 3,
+				"output": 15,
+				"cacheRead": 0.3,
+				"cacheWrite": 3.75
+			},
+			"contextWindow": 200000,
+			"maxTokens": 64000,
+			"thinking": {
+				"mode": "effort",
+				"minLevel": "minimal",
+				"maxLevel": "xhigh"
+			}
+		},
+		"google/gemini-2.5-pro": {
+			"id": "google/gemini-2.5-pro",
+			"name": "Gemini 2.5 Pro",
+			"api": "openai-completions",
+			"provider": "bizrouter",
+			"baseUrl": "https://api.bizrouter.ai/v1",
+			"reasoning": true,
+			"input": [
+				"text",
+				"image"
+			],
+			"cost": {
+				"input": 1.25,
+				"output": 10,
+				"cacheRead": 0.31,
+				"cacheWrite": 0
+			},
+			"contextWindow": 1048576,
+			"maxTokens": 65536,
+			"thinking": {
+				"mode": "effort",
+				"minLevel": "minimal",
+				"maxLevel": "high"
+			}
+		},
+		"openai/gpt-4o": {
+			"id": "openai/gpt-4o",
+			"name": "GPT-4o",
+			"api": "openai-completions",
+			"provider": "bizrouter",
+			"baseUrl": "https://api.bizrouter.ai/v1",
+			"reasoning": false,
+			"input": [
+				"text",
+				"image"
+			],
+			"cost": {
+				"input": 2.5,
+				"output": 10,
+				"cacheRead": 1.25,
+				"cacheWrite": 0
+			},
+			"contextWindow": 128000,
+			"maxTokens": 16384
+		}
+	},
 	"cerebras": {
 		"gemma-4-31b": {
 			"id": "gemma-4-31b",

package/src/provider-models/descriptors.ts

@@ -11,6 +11,7 @@ import { ollamaCloudModelManagerOptions } from "./ollama";
 import {
 	alibabaTokenPlanModelManagerOptions,
 	anthropicModelManagerOptions,
+	bizrouterModelManagerOptions,
 	cerebrasModelManagerOptions,
 	cloudflareAiGatewayModelManagerOptions,
 	deepinfraModelManagerOptions,
@@ -319,6 +320,12 @@ export const PROVIDER_DESCRIPTORS: readonly ProviderDescriptor[] = [
 		config => opengatewayModelManagerOptions(config),
 		catalog("OpenGateway by Sionic AI", ["OPENGATEWAY_API_KEY"]),
 	),
+	catalogDescriptor(
+		"bizrouter",
+		"anthropic/claude-sonnet-4.5",
+		config => bizrouterModelManagerOptions(config),
+		catalog("BizRouter", ["BIZROUTER_API_KEY"]),
+	),
 	catalogDescriptor("zai", "glm-5.2", config => zaiModelManagerOptions(config), catalog("zAI", ["ZAI_API_KEY"])),
 	catalogDescriptor(
 		"glm-zcode",