Skip to content

Never stop coding. Free MIT AI gateway: one endpoint, 375 providers (150+ free), 1,400+ models Kimi, Claude, GPT, Gemini, GLM, DeepSeek, MiniMax. Works with Claude Code, Codex, Cursor, OpenCode, Cline & Copilot. Quota-aware auto-fallback, RTK+Caveman compression saves 15-95% tokens, MCP/A2A, Desktop/PWA. Built by 900+ contributors

59.6 KB
Raw Blame
import test from "node:test";
import assert from "node:assert/strict";

import {
  applyConfiguredUserAgent,
  BaseExecutor,
  getCustomUserAgent,
  mergeAbortSignals,
  mergeUpstreamExtraHeaders,
  setUserAgentHeader,
} from "../../open-sse/executors/base.ts";
import { shouldForceResponsesUpstream } from "../../open-sse/executors/forceResponsesUpstream.ts";
import { DefaultExecutor } from "../../open-sse/executors/default.ts";
import { NO_REFRESH_NEEDED } from "../../open-sse/services/tokenRefresh/shared.ts";
import { PROVIDERS } from "../../open-sse/config/constants.ts";
import {
  CLAUDE_CODE_COMPATIBLE_ANTHROPIC_VERSION,
  CLAUDE_CODE_COMPATIBLE_DEFAULT_CHAT_PATH,
  CLAUDE_CODE_COMPATIBLE_REDACT_THINKING_BETA,
  CONTEXT_1M_BETA_HEADER,
} from "../../open-sse/services/claudeCodeCompatible.ts";
import { runWithCapture } from "../../open-sse/utils/providerRequestLogging.ts";
import { CLAUDE_CODE_CLIENT_BILLING_VERSION } from "../../src/shared/constants/claudeCodeClient.ts";

class TestExecutor extends BaseExecutor {
  constructor(config = {}) {
    super("test-provider", {
      baseUrls: [
        "https://primary.example/v1/chat/completions",
        "https://fallback.example/v1/chat/completions",
      ],
      headers: { "X-Test-Header": "base" },
      ...config,
    });
  }

  async transformRequest(model, body, stream) {
    return { ...body, transformed: true, model, stream };
  }
}

test("BaseExecutor: openai-compatible buildUrl sanitizes custom chat paths", () => {
  const executor = new BaseExecutor("openai-compatible-test", {});
  const valid = executor.buildUrl("gpt-4.1", true, 0, {
    providerSpecificData: {
      baseUrl: "https://proxy.example/v1/",
      chatPath: "/custom/chat/completions",
    },
  });
  const invalid = executor.buildUrl("gpt-4.1", true, 0, {
    providerSpecificData: {
      baseUrl: "https://proxy.example/v1/",
      chatPath: "../evil",
    },
  });
  const invalidNullByte = executor.buildUrl("gpt-4.1", true, 0, {
    providerSpecificData: {
      baseUrl: "https://proxy.example/v1/",
      chatPath: "/ok\0evil",
    },
  });

  assert.equal(valid, "https://proxy.example/v1/custom/chat/completions");
  assert.equal(invalid, "https://proxy.example/v1/chat/completions");
  assert.equal(invalidNullByte, "https://proxy.example/v1/chat/completions");
});

test("BaseExecutor: legacy openai-compatible providers honor providerSpecificData.apiType", () => {
  const executor = new BaseExecutor("openai-compatible-sp-openai", {});
  const url = executor.buildUrl("gpt-5.4", true, 0, {
    providerSpecificData: {
      apiType: "responses",
      baseUrl: "https://proxy.example/v1/",
    },
  });

  assert.equal(url, "https://proxy.example/v1/responses");
});

test("DefaultExecutor.buildUrl handles Gemini and Claude variants", () => {
  const gemini = new DefaultExecutor("gemini");
  const claude = new DefaultExecutor("claude");

  assert.equal(
    gemini.buildUrl("gemini-2.5-flash", false),
    "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:generateContent"
  );
  assert.equal(
    gemini.buildUrl("gemini-2.5-flash", true),
    "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:streamGenerateContent?alt=sse"
  );
  assert.equal(claude.buildUrl("claude-sonnet-4", true), `${PROVIDERS.claude.baseUrl}?beta=true`);
});

test("DefaultExecutor.buildUrl uses full chat endpoints for hosted OpenAI-compatible providers", () => {
  const bazaarlink = new DefaultExecutor("bazaarlink");
  const crof = new DefaultExecutor("crof");

  assert.equal(
    bazaarlink.buildUrl("auto:free", true),
    "https://bazaarlink.ai/api/v1/chat/completions"
  );
  assert.equal(crof.buildUrl("gpt-4.1", true), "https://crof.ai/v1/chat/completions");
});

test("DefaultExecutor.buildUrl honors a custom providerSpecificData.baseUrl for the built-in openai provider", () => {
  const openai = new DefaultExecutor("openai");

  // No override → hardcoded OpenAI endpoint (unchanged behavior).
  assert.equal(openai.buildUrl("gpt-4o", true), "https://api.openai.com/v1/chat/completions");

  // Custom base URL (e.g. a proxy/gateway) must be used instead of api.openai.com.
  assert.equal(
    openai.buildUrl("gpt-4o", true, 0, {
      providerSpecificData: { baseUrl: "https://api.contactboxtools.me/v1" },
    }),
    "https://api.contactboxtools.me/v1/chat/completions"
  );

  // Trailing slash is normalized.
  assert.equal(
    openai.buildUrl("gpt-4o", true, 0, {
      providerSpecificData: { baseUrl: "https://proxy.example/v1/" },
    }),
    "https://proxy.example/v1/chat/completions"
  );

  // A base URL already pointing at the chat endpoint is kept as-is.
  assert.equal(
    openai.buildUrl("gpt-4o", true, 0, {
      providerSpecificData: { baseUrl: "https://proxy.example/v1/chat/completions" },
    }),
    "https://proxy.example/v1/chat/completions"
  );
});

test("DefaultExecutor.buildUrl handles openai-compatible and anthropic-compatible providers", () => {
  const openAICompat = new DefaultExecutor("openai-compatible-test");
  const openAIResponsesCompat = new DefaultExecutor("openai-compatible-responses-test");
  const openAILegacyResponsesCompat = new DefaultExecutor("openai-compatible-sp-openai");
  const anthropicCompat = new DefaultExecutor("anthropic-compatible-test");
  const anthropicCcCompat = new DefaultExecutor("anthropic-compatible-cc-test");

  assert.equal(
    openAICompat.buildUrl("gpt-4.1", true, 0, {
      providerSpecificData: { baseUrl: "https://proxy.example/v1/" },
    }),
    "https://proxy.example/v1/chat/completions"
  );
  assert.equal(
    openAICompat.buildUrl("gpt-4.1", true, 0, {
      providerSpecificData: {
        baseUrl: "https://proxy.example/v1/",
        chatPath: "/custom/chat",
      },
    }),
    "https://proxy.example/v1/custom/chat"
  );
  assert.equal(
    openAIResponsesCompat.buildUrl("gpt-4.1", true, 0, {
      providerSpecificData: { baseUrl: "https://proxy.example/v1/" },
    }),
    "https://proxy.example/v1/responses"
  );
  assert.equal(
    openAICompat.buildUrl("gpt-4.1", true, 0, {
      providerSpecificData: {
        baseUrl: "https://proxy.example/v1/",
        _omnirouteForceResponsesUpstream: true,
      },
    }),
    "https://proxy.example/v1/responses"
  );
  assert.equal(
    openAILegacyResponsesCompat.buildUrl("gpt-5.4", true, 0, {
      providerSpecificData: {
        apiType: "responses",
        baseUrl: "https://proxy.example/v1/",
      },
    }),
    "https://proxy.example/v1/responses"
  );
  assert.equal(
    anthropicCompat.buildUrl("claude-sonnet-4", true, 0, {
      providerSpecificData: { baseUrl: "https://anthropic.example/v1/" },
    }),
    "https://anthropic.example/v1/messages"
  );
  assert.equal(
    anthropicCompat.buildUrl("claude-sonnet-4", true, 0, {
      providerSpecificData: {
        baseUrl: "https://anthropic.example/v1/",
        chatPath: "/custom/messages",
      },
    }),
    "https://anthropic.example/v1/custom/messages"
  );
  assert.equal(
    anthropicCcCompat.buildUrl("claude-sonnet-4", true, 0, {
      providerSpecificData: {
        baseUrl: "https://cc.example/v1/messages",
      },
    }),
    `https://cc.example${CLAUDE_CODE_COMPATIBLE_DEFAULT_CHAT_PATH}`
  );
});

test("DefaultExecutor.buildUrl normalizes configurable chat-openai-compat base URLs", () => {
  const bailian = new DefaultExecutor("bailian-coding-plan");
  const heroku = new DefaultExecutor("heroku");
  const databricks = new DefaultExecutor("databricks");
  const azureAi = new DefaultExecutor("azure-ai");
  const watsonx = new DefaultExecutor("watsonx");
  const oci = new DefaultExecutor("oci");
  const sap = new DefaultExecutor("sap");
  const modal = new DefaultExecutor("modal");
  const reka = new DefaultExecutor("reka");
  const maritalk = new DefaultExecutor("maritalk");
  const snowflake = new DefaultExecutor("snowflake");
  const gigachat = new DefaultExecutor("gigachat");
  const siliconflow = new DefaultExecutor("siliconflow");

  assert.equal(
    bailian.buildUrl("qwen3-coder-plus", true, 0, {
      providerSpecificData: {
        baseUrl: "https://token-plan.ap-southeast-1.maas.aliyuncs.com/apps/anthropic/v1",
      },
    }),
    "https://token-plan.ap-southeast-1.maas.aliyuncs.com/apps/anthropic/v1/messages"
  );
  assert.equal(
    heroku.buildUrl("claude-4-sonnet", true, 0, {
      providerSpecificData: { baseUrl: "https://us.inference.heroku.com" },
    }),
    "https://us.inference.heroku.com/v1/chat/completions"
  );
  assert.equal(
    databricks.buildUrl("databricks-gpt-5", true, 0, {
      providerSpecificData: {
        baseUrl: "https://adb-1234567890123456.7.azuredatabricks.net/serving-endpoints",
      },
    }),
    "https://adb-1234567890123456.7.azuredatabricks.net/serving-endpoints/chat/completions"
  );

  assert.equal(
    azureAi.buildUrl("DeepSeek-V3.1", true, 0, {
      providerSpecificData: { baseUrl: "https://my-foundry.services.ai.azure.com" },
    }),
    "https://my-foundry.services.ai.azure.com/openai/v1/chat/completions"
  );

  assert.equal(
    watsonx.buildUrl("ibm/granite-3-3-8b-instruct", true, 0, {
      providerSpecificData: { baseUrl: "https://ca-tor.ml.cloud.ibm.com" },
    }),
    "https://ca-tor.ml.cloud.ibm.com/ml/gateway/v1/chat/completions"
  );
  assert.equal(
    oci.buildUrl("openai.gpt-oss-20b", true, 0, {
      providerSpecificData: {
        baseUrl: "https://inference.generativeai.us-ashburn-1.oci.oraclecloud.com",
      },
    }),
    "https://inference.generativeai.us-ashburn-1.oci.oraclecloud.com/openai/v1/chat/completions"
  );
  assert.equal(
    sap.buildUrl("gpt-4o", true, 0, {
      providerSpecificData: {
        baseUrl: "https://sap.example.com/v2/lm/deployments/demo-deployment",
      },
    }),
    "https://sap.example.com/v2/lm/deployments/demo-deployment/chat/completions"
  );
  assert.equal(
    modal.buildUrl("Qwen/Qwen3-4B-Thinking-2507-FP8", true, 0, {
      providerSpecificData: {
        baseUrl: "https://alice--demo.modal.run/v1",
      },
    }),
    "https://alice--demo.modal.run/v1/chat/completions"
  );
  assert.equal(
    reka.buildUrl("reka-core", true, 0, {
      providerSpecificData: {
        baseUrl: "https://api.reka.ai/v1",
      },
    }),
    "https://api.reka.ai/v1/chat/completions"
  );
  assert.equal(
    maritalk.buildUrl("sabia-4", true, 0, {
      providerSpecificData: {
        baseUrl: "https://chat.maritaca.ai/api/chat/inference",
      },
    }),
    "https://chat.maritaca.ai/api/chat/completions"
  );
  assert.equal(
    snowflake.buildUrl("llama3.3-70b", true, 0, {
      providerSpecificData: { baseUrl: "https://account.snowflakecomputing.com" },
    }),
    "https://account.snowflakecomputing.com/api/v2/cortex/inference:complete"
  );
  assert.equal(
    gigachat.buildUrl("GigaChat-2-Pro", true, 0, {
      providerSpecificData: { baseUrl: "https://gigachat.devices.sberbank.ru/api/v1" },
    }),
    "https://gigachat.devices.sberbank.ru/api/v1/chat/completions"
  );
  assert.equal(
    siliconflow.buildUrl("deepseek-ai/DeepSeek-V3.2", true),
    "https://api.siliconflow.com/v1/chat/completions"
  );
  assert.equal(
    siliconflow.buildUrl("deepseek-ai/DeepSeek-V3.2", true, 0, {
      providerSpecificData: { baseUrl: "https://api.siliconflow.cn/v1" },
    }),
    "https://api.siliconflow.cn/v1/chat/completions"
  );
});

test("DefaultExecutor.buildUrl falls back to OpenAI config for unknown providers", () => {
  const executor = new DefaultExecutor("unknown-provider");
  assert.equal(executor.config.baseUrl, PROVIDERS.openai.baseUrl);
  assert.equal(executor.buildUrl("gpt-4.1", true), PROVIDERS.openai.baseUrl);
});

test("DefaultExecutor.buildUrl applies urlSuffix for zai and glm-coding-apikey", () => {
  const zai = new DefaultExecutor("zai");
  const glmCodingApikey = new DefaultExecutor("glm-coding-apikey");
  assert.equal(
    zai.buildUrl("glm-5", true, 0, {
      providerSpecificData: { baseUrl: "https://api.z.ai/api/anthropic/v1/messages" },
    }),
    "https://api.z.ai/api/anthropic/v1/messages?beta=true"
  );
  assert.equal(
    glmCodingApikey.buildUrl("glm-4.7", true, 0, {
      providerSpecificData: { baseUrl: "https://api.z.ai/api/anthropic/v1/messages" },
    }),
    "https://api.z.ai/api/anthropic/v1/messages?beta=true"
  );
  assert.equal(zai.buildUrl("glm-5", true), "https://api.z.ai/api/anthropic/v1/messages?beta=true");
});

test("DefaultExecutor.buildUrl applies urlSuffix from registry for unknown providers with suffix", () => {
  const executor = new DefaultExecutor("unknown-provider");
  assert.equal(executor.buildUrl("gpt-4.1", true), PROVIDERS.openai.baseUrl);
});

test("DefaultExecutor.buildHeaders uses x-api-key for zai and glm-coding-apikey", () => {
  const zai = new DefaultExecutor("zai");
  const glmCodingApikey = new DefaultExecutor("glm-coding-apikey");
  const zaiHeaders = zai.buildHeaders({ apiKey: "zai-key" }, true);
  const glmHeaders = glmCodingApikey.buildHeaders({ apiKey: "glm-key" }, true);
  assert.equal(zaiHeaders["x-api-key"], "zai-key");
  assert.equal(glmHeaders["x-api-key"], "glm-key");
  assert.equal(zaiHeaders["Authorization"], undefined);
  assert.equal(glmHeaders["Authorization"], undefined);
});

test("DefaultExecutor.buildHeaders handles Gemini and Claude auth modes", () => {
  const gemini = new DefaultExecutor("gemini");
  const claude = new DefaultExecutor("claude");
  const azureAi = new DefaultExecutor("azure-ai");
  const oci = new DefaultExecutor("oci");
  const sap = new DefaultExecutor("sap");
  const modal = new DefaultExecutor("modal");
  const maritalk = new DefaultExecutor("maritalk");

  const geminiApiKeyHeaders = gemini.buildHeaders({ apiKey: "gem-key" }, true);
  const geminiOAuthHeaders = gemini.buildHeaders({ accessToken: "gem-token" }, false);
  const claudeApiKeyHeaders = claude.buildHeaders({ apiKey: "claude-key" }, true);
  const claudeOAuthHeaders = claude.buildHeaders({ accessToken: "claude-token" }, false);
  const azureAiHeaders = azureAi.buildHeaders({ apiKey: "azure-ai-key" }, true);
  const ociHeaders = oci.buildHeaders(
    {
      apiKey: "oci-key",
      projectId: "ocid1.generativeaiproject.oc1.us-chicago-1.example",
    },
    true
  );
  const sapHeaders = sap.buildHeaders(
    {
      apiKey: "sap-key",
      providerSpecificData: {
        resourceGroup: "shared",
      },
    },
    true
  );
  const modalHeaders = modal.buildHeaders(
    {
      apiKey: "modal-key",
    },
    true
  );
  const maritalkHeaders = maritalk.buildHeaders({ apiKey: "maritalk-key" }, true);

  assert.equal(geminiApiKeyHeaders["x-goog-api-key"], "gem-key");
  assert.equal(geminiApiKeyHeaders.Accept, "text/event-stream");
  assert.equal(geminiApiKeyHeaders.Authorization, undefined);
  assert.equal(geminiOAuthHeaders.Authorization, "Bearer gem-token");
  assert.equal(claudeApiKeyHeaders["x-api-key"], "claude-key");
  assert.equal(claudeApiKeyHeaders.Accept, "text/event-stream");
  assert.equal(claudeOAuthHeaders.Authorization, "Bearer claude-token");
  assert.equal(claudeOAuthHeaders["x-api-key"], undefined);
  assert.equal(azureAiHeaders["api-key"], "azure-ai-key");
  assert.equal(azureAiHeaders.Authorization, undefined);
  assert.equal(ociHeaders.Authorization, "Bearer oci-key");
  assert.equal(ociHeaders["OpenAI-Project"], "ocid1.generativeaiproject.oc1.us-chicago-1.example");
  assert.equal(sapHeaders.Authorization, "Bearer sap-key");
  assert.equal(sapHeaders["AI-Resource-Group"], "shared");
  assert.equal(modalHeaders.Authorization, "Bearer modal-key");
  assert.equal(maritalkHeaders.Authorization, "Key maritalk-key");
});

test("DefaultExecutor.buildHeaders handles GLM, default auth and anthropic-compatible headers", () => {
  const glm = new DefaultExecutor("glm");
  const glmt = new DefaultExecutor("glmt");
  const openai = new DefaultExecutor("openai");
  const anthropicCompat = new DefaultExecutor("anthropic-compatible-test");

  const glmHeaders = glm.buildHeaders({ accessToken: "glm-token" }, false);
  const glmtHeaders = glmt.buildHeaders({ apiKey: "glmt-key" }, false);
  const openaiHeaders = openai.buildHeaders({ apiKey: "sk-openai" }, true);
  const anthropicHeaders = anthropicCompat.buildHeaders({ apiKey: "anth-key" }, true);

  assert.equal(glmHeaders["x-api-key"], "glm-token");
  assert.equal(glmtHeaders["x-api-key"], "glmt-key");
  assert.equal(openaiHeaders.Authorization, "Bearer sk-openai");
  assert.equal(openaiHeaders.Accept, "text/event-stream");
  assert.equal(anthropicHeaders["x-api-key"], "anth-key");
  assert.equal(anthropicHeaders["anthropic-version"], "2023-06-01");
  assert.equal(anthropicHeaders.Accept, "text/event-stream");
});

test("DefaultExecutor.buildHeaders keeps a caller-supplied Anthropic-Version (case-insensitive guard) for anthropic-compatible providers", () => {
  // An operator may configure a Title-Case "Anthropic-Version" via the provider
  // config headers. The default-guard at the anthropic-compatible-* branch must
  // detect it case-insensitively and NOT add a second lowercase
  // "anthropic-version" key, which undici would otherwise combine into
  // "2025-01-01, 2023-06-01" and break the upstream request.
  const anthropicCompat = new DefaultExecutor("anthropic-compatible-test");
  // `config` is shared across instances via the provider registry, so snapshot
  // and restore `config.headers` to avoid leaking the Title-Case override into
  // other tests.
  const originalConfigHeaders = anthropicCompat.config.headers;
  anthropicCompat.config.headers = {
    ...originalConfigHeaders,
    "Anthropic-Version": "2025-01-01",
  };

  try {
    const headers = anthropicCompat.buildHeaders({ apiKey: "anth-key" }, true);

    const versionKeys = Object.keys(headers).filter(
      (key) => key.toLowerCase() === "anthropic-version"
    );
    assert.equal(versionKeys.length, 1, "Duplicate anthropic-version header keys found");
    assert.equal(headers["Anthropic-Version"], "2025-01-01");
    assert.equal(headers["anthropic-version"], undefined);
    assert.equal(headers["x-api-key"], "anth-key");
  } finally {
    anthropicCompat.config.headers = originalConfigHeaders;
  }
});

test("DefaultExecutor.buildHeaders still defaults anthropic-version when no variant is present", () => {
  const anthropicCompat = new DefaultExecutor("anthropic-compatible-test");
  const headers = anthropicCompat.buildHeaders({ apiKey: "anth-key" }, true);
  assert.equal(headers["anthropic-version"], "2023-06-01");
});

test("DefaultExecutor local OpenAI-style providers honor custom base URLs and skip empty bearer headers", () => {
  const lmStudio = new DefaultExecutor("lm-studio");
  const vllm = new DefaultExecutor("vllm");

  const lmStudioUrl = lmStudio.buildUrl("local-model", true, 0, {
    providerSpecificData: { baseUrl: "http://127.0.0.1:4321/v1" },
  });
  const vllmHeaders = vllm.buildHeaders({}, false);

  assert.equal(lmStudioUrl, "http://127.0.0.1:4321/v1/chat/completions");
  assert.equal(vllmHeaders.Authorization, undefined);
  assert.equal(vllmHeaders.Accept, "application/json");
});

test("DefaultExecutor local providers append /v1/chat/completions for bare hostname base URLs", () => {
  const llamaCpp = new DefaultExecutor("llama-cpp");
  const lmStudio = new DefaultExecutor("lm-studio");
  const vllm = new DefaultExecutor("vllm");

  const bareHost = llamaCpp.buildUrl("gemma-4", true, 0, {
    providerSpecificData: { baseUrl: "https://foo.llama.example.com" },
  });
  const customPath = lmStudio.buildUrl("gemma-4", true, 0, {
    providerSpecificData: { baseUrl: "https://bar.llama.ai/foo" },
  });
  const alreadyComplete = vllm.buildUrl("gemma-4", true, 0, {
    providerSpecificData: { baseUrl: "https://baz.llama.ai/v1/chat/completions" },
  });

  assert.equal(bareHost, "https://foo.llama.example.com/v1/chat/completions");
  assert.equal(customPath, "https://bar.llama.ai/foo/v1/chat/completions");
  assert.equal(alreadyComplete, "https://baz.llama.ai/v1/chat/completions");
});

test("DefaultExecutor.buildHeaders handles Snowflake PATs and GigaChat access tokens", () => {
  const snowflake = new DefaultExecutor("snowflake");
  const gigachat = new DefaultExecutor("gigachat");

  const snowflakePatHeaders = snowflake.buildHeaders({ apiKey: "pat/test-token" }, false);
  const snowflakeJwtHeaders = snowflake.buildHeaders({ apiKey: "jwt-token" }, false);
  const gigachatHeaders = gigachat.buildHeaders({ accessToken: "gigachat-token" }, false);

  assert.equal(snowflakePatHeaders.Authorization, "Bearer test-token");
  assert.equal(
    snowflakePatHeaders["X-Snowflake-Authorization-Token-Type"],
    "PROGRAMMATIC_ACCESS_TOKEN"
  );
  assert.equal(snowflakeJwtHeaders.Authorization, "Bearer jwt-token");
  assert.equal(snowflakeJwtHeaders["X-Snowflake-Authorization-Token-Type"], "KEYPAIR_JWT");
  assert.equal(gigachatHeaders.Authorization, "Bearer gigachat-token");
});

test("DefaultExecutor.buildHeaders rotates extra API keys and builds Claude Code compatible headers", () => {
  const openai = new DefaultExecutor("openai");
  const cc = new DefaultExecutor("anthropic-compatible-cc-test");

  const first = openai.buildHeaders(
    {
      apiKey: "primary",
      connectionId: "conn-rotation",
      providerSpecificData: { extraApiKeys: ["extra-1", "extra-2"] },
    },
    false
  );
  const second = openai.buildHeaders(
    {
      apiKey: "primary",
      connectionId: "conn-rotation",
      providerSpecificData: { extraApiKeys: ["extra-1", "extra-2"] },
    },
    false
  );
  const ccHeaders = cc.buildHeaders(
    {
      apiKey: "cc-key",
      providerSpecificData: { ccSessionId: "session-1" },
    },
    true
  );
  const ccJsonHeaders = cc.buildHeaders(
    {
      apiKey: "cc-key",
      providerSpecificData: { ccSessionId: "session-1" },
    },
    false
  );

  assert.equal(first.Authorization, "Bearer primary");
  assert.equal(second.Authorization, "Bearer extra-1");
  assert.equal(ccHeaders.Authorization, "Bearer cc-key");
  assert.equal(ccHeaders["x-api-key"], undefined);
  assert.equal(ccHeaders["anthropic-version"], CLAUDE_CODE_COMPATIBLE_ANTHROPIC_VERSION);
  assert.equal(ccHeaders["X-Claude-Code-Session-Id"], "session-1");
  assert.equal(ccHeaders.Accept, "text/event-stream");
  assert.equal(ccJsonHeaders.Accept, "application/json");
});

test("DefaultExecutor.execute uses CC-compatible connection defaults to append 1M beta", async () => {
  const originalFetch = globalThis.fetch;
  const calls = [];
  const toPlainHeaders = (headers) =>
    headers instanceof Headers
      ? Object.fromEntries(headers.entries())
      : Object.fromEntries(
          Object.entries(headers || {}).map(([key, value]) => [
            key,
            value == null ? "" : String(value),
          ])
        );

  globalThis.fetch = async (_url, init = {}) => {
    calls.push({ headers: toPlainHeaders(init.headers) });
    return new Response(JSON.stringify({ ok: true }), {
      status: 200,
      headers: { "Content-Type": "application/json" },
    });
  };

  try {
    const cc = new DefaultExecutor("anthropic-compatible-cc-test");
    await cc.execute({
      model: "claude-sonnet-4-6",
      body: {
        model: "claude-sonnet-4-6",
        messages: [{ role: "user", content: "hi" }],
        max_tokens: 1,
      },
      stream: false,
      credentials: {
        apiKey: "cc-key",
        // #13452: buildUrl() now requires a hydrated baseUrl.
        providerSpecificData: { ccSessionId: "session-1", baseUrl: "https://cc.test/v1" },
      },
      clientHeaders: {
        "x-app": "cli",
        "user-agent": "claude-cli/2.1.116 (external, cli)",
      },
      extendedContext: false,
    });
    await cc.execute({
      model: "claude-sonnet-4-6",
      body: {
        model: "claude-sonnet-4-6",
        messages: [{ role: "user", content: "hi" }],
        max_tokens: 1,
      },
      stream: false,
      credentials: {
        apiKey: "cc-key",
        providerSpecificData: {
          ccSessionId: "session-1",
          baseUrl: "https://cc.test/v1",
          requestDefaults: { context1m: true, redactThinking: true },
        },
      },
      extendedContext: false,
    });

    const anthropicCompat = new DefaultExecutor("anthropic-compatible-test");
    await anthropicCompat.execute({
      model: "claude-sonnet-4-6",
      body: {
        model: "claude-sonnet-4-6",
        messages: [{ role: "user", content: "hi" }],
        max_tokens: 1,
      },
      stream: false,
      credentials: {
        apiKey: "anth-key",
        providerSpecificData: {
          baseUrl: "https://anthropic.example.com/v1",
        },
      },
      extendedContext: true,
    });
    await cc.execute({
      model: "claude-opus-5",
      body: {
        model: "claude-opus-5",
        messages: [{ role: "user", content: "hi" }],
        max_tokens: 1,
      },
      stream: false,
      credentials: {
        apiKey: "cc-key",
        providerSpecificData: {
          ccSessionId: "session-1",
          baseUrl: "https://cc-proxy.example.test/v1",
          requestDefaults: { context1m: true },
        },
      },
      extendedContext: true,
    });
  } finally {
    globalThis.fetch = originalFetch;
  }

  assert.equal(calls[0].headers["anthropic-beta"].includes(CONTEXT_1M_BETA_HEADER), false);
  assert.equal(
    calls[0].headers["anthropic-beta"].includes(CLAUDE_CODE_COMPATIBLE_REDACT_THINKING_BETA),
    false
  );
  assert.equal(calls[1].headers["anthropic-beta"].includes(CONTEXT_1M_BETA_HEADER), true);
  assert.equal(
    calls[1].headers["anthropic-beta"].includes(CLAUDE_CODE_COMPATIBLE_REDACT_THINKING_BETA),
    true
  );
  // claude-sonnet-4-6 GA'd 1M context (2026-02-17) and was added to CONTEXT_1M_SUPPORTED_MODELS
  // by #7129; a non-CC anthropic-compatible target with extendedContext:true now legitimately
  // gets the context-1m beta header (shouldForwardExtendedContext in base.ts), same as any other
  // 1M-capable model.
  assert.equal(calls[2].headers["anthropic-beta"].includes(CONTEXT_1M_BETA_HEADER), true);
  // Opus 5 has a native 1M window and must not receive the legacy context beta.
  assert.equal(calls[3].headers["anthropic-beta"].includes(CONTEXT_1M_BETA_HEADER), false);
});

test("DefaultExecutor.execute reports the exact serialized provider request before fetch", async () => {
  const originalFetch = globalThis.fetch;
  let fetchStarted = false;
  let fetchBody: any = null;
  let prepared: any = null;
  let preparedBeforeFetch = false;

  globalThis.fetch = async (_url, init = {}) => {
    fetchStarted = true;
    fetchBody = JSON.parse(String(init.body || "{}"));
    return new Response(JSON.stringify({ ok: true }), {
      status: 200,
      headers: { "Content-Type": "application/json" },
    });
  };

  try {
    const cc = new DefaultExecutor("anthropic-compatible-cc-test");
    const requestCapture = {
      capture(request) {
        preparedBeforeFetch = !fetchStarted;
        prepared = request;
      },
      body(fallback) {
        return prepared?.body ?? fallback;
      },
      latest() {
        return prepared;
      },
    };
    const result = await runWithCapture(requestCapture, () =>
      cc.execute({
        model: "claude-sonnet-4-6",
        body: {
          model: "claude-sonnet-4-6",
          system: [
            {
              type: "text",
              text: "x-anthropic-billing-header: cc_version=1.0.0; cc_entrypoint=sdk-cli; cch=00000;",
            },
          ],
          messages: [{ role: "user", content: "hi" }],
          max_tokens: 1,
          reasoning_effort: "xhigh",
        },
        stream: false,
        credentials: {
          apiKey: "cc-key",
          providerSpecificData: { ccSessionId: "session-1", baseUrl: "https://cc.test/v1" }, // #13452
        },
      })
    );

    assert.ok(prepared, "prepared request hook should fire before fetch");
    assert.equal(preparedBeforeFetch, true);
    assert.deepEqual(prepared.body, fetchBody);
    assert.deepEqual(result.transformedBody, fetchBody);
    assert.equal(prepared.body.reasoning_effort, "max");
    assert.equal(fetchBody.reasoning_effort, "max");
    assert.match(JSON.stringify(fetchBody), /\bcch=(?!00000)[0-9a-f]{5};/);
  } finally {
    globalThis.fetch = originalFetch;
  }
});

test("DefaultExecutor.execute only injects adaptive thinking defaults for Claude models that support x-high effort", async () => {
  const originalFetch = globalThis.fetch;
  const requestBodies = [];

  globalThis.fetch = async (_url, init = {}) => {
    requestBodies.push(JSON.parse(String(init.body)));
    return new Response(JSON.stringify({ ok: true }), {
      status: 200,
      headers: { "Content-Type": "application/json" },
    });
  };

  try {
    const claude = new DefaultExecutor("claude");
    await claude.execute({
      model: "claude-opus-4-7",
      body: {
        model: "claude-opus-4-7",
        messages: [{ role: "user", content: "hi" }],
        max_tokens: 1,
      },
      stream: false,
      credentials: {
        apiKey: "cc-key",
        providerSpecificData: {
          ccSessionId: "session-1",
        },
      },
      clientHeaders: {
        "x-app": "cli",
        "user-agent": "claude-cli/2.1.116 (external, cli)",
      },
      extendedContext: false,
    });

    await claude.execute({
      model: "claude-haiku-4-5-20251001",
      body: {
        model: "claude-haiku-4-5-20251001",
        messages: [{ role: "user", content: "hi" }],
        max_tokens: 1,
      },
      stream: false,
      credentials: {
        apiKey: "cc-key",
        providerSpecificData: {
          ccSessionId: "session-1",
        },
      },
      clientHeaders: {
        "x-app": "cli",
        "user-agent": "claude-cli/2.1.116 (external, cli)",
      },
      extendedContext: false,
    });

    await claude.execute({
      model: "claude-sonnet-4-6",
      body: {
        model: "claude-sonnet-4-6",
        messages: [{ role: "user", content: "hi" }],
        max_tokens: 1,
        thinking: { type: "disabled" },
      },
      stream: false,
      credentials: {
        apiKey: "cc-key",
        providerSpecificData: {
          ccSessionId: "session-1",
        },
      },
      clientHeaders: {
        "x-app": "cli",
        "user-agent": "claude-cli/2.1.116 (external, cli)",
      },
      extendedContext: false,
    });
  } finally {
    globalThis.fetch = originalFetch;
  }

  assert.deepEqual((requestBodies[0] as any).thinking, { type: "adaptive" });
  assert.deepEqual((requestBodies[0] as any).context_management, {
    edits: [{ type: "clear_thinking_20251015", keep: "all" }],
  });
  assert.deepEqual((requestBodies[0] as any).output_config, { effort: "high" });

  assert.equal((requestBodies[1] as any).thinking, undefined);
  assert.equal((requestBodies[1] as any).context_management, undefined);
  assert.equal((requestBodies[1] as any).output_config, undefined);

  assert.deepEqual((requestBodies[2] as any).thinking, { type: "disabled" });
  assert.equal((requestBodies[2] as any).context_management, undefined);
});

test("DefaultExecutor.transformRequest injects OpenAI stream usage and preserves model ids with slashes", () => {
  const executor = new DefaultExecutor("openai");
  const body = { model: "zai-org/GLM-5-FP8", messages: [{ role: "user", content: "hi" }] };
  const result = executor.transformRequest("zai-org/GLM-5-FP8", body, true, {});

  assert.notEqual(result, body);
  assert.equal(result.model, "zai-org/GLM-5-FP8");
  assert.deepEqual((result as any).stream_options, { include_usage: true });
  assert.equal((body as any).stream_options, undefined);
});

test("DefaultExecutor.transformRequest only injects stream usage for OpenAI chat targets", () => {
  const openAICompat = new DefaultExecutor("openai-compatible-test");
  const openAIResponsesCompat = new DefaultExecutor("openai-compatible-responses-test");

  const chatBody = { model: "gpt-4.1", messages: [{ role: "user", content: "hi" }] };
  const responsesBody = { model: "gpt-4.1", input: "hi" };

  const chatResult = openAICompat.transformRequest("gpt-4.1", chatBody, true, {
    providerSpecificData: { baseUrl: "https://proxy.example/v1" },
  });
  const responsesResult = openAIResponsesCompat.transformRequest("gpt-4.1", responsesBody, true, {
    providerSpecificData: { baseUrl: "https://proxy.example/v1" },
  });

  assert.deepEqual((chatResult as any).stream_options, { include_usage: true });
  assert.equal((responsesResult as any).stream_options, undefined);
});

test("DefaultExecutor.execute routes Responses-shaped MCP requests to /responses for OpenAI-compatible providers", async () => {
  const originalFetch = globalThis.fetch;
  const calls: Array<{ url: string; body: any }> = [];

  globalThis.fetch = async (url, init = {}) => {
    calls.push({
      url: String(url),
      body: JSON.parse(String(init.body)),
    });
    return new Response(JSON.stringify({ ok: true }), {
      status: 200,
      headers: { "Content-Type": "application/json" },
    });
  };

  try {
    const executor = new DefaultExecutor("openai-compatible-test");
    await executor.execute({
      model: "gpt-4.1",
      body: {
        model: "gpt-4.1",
        input: "find tools",
        tools: [{ type: "tool_search" }],
      },
      stream: false,
      credentials: {
        apiKey: "test-key",
        providerSpecificData: { baseUrl: "https://proxy.example/v1/" },
      },
    });
  } finally {
    globalThis.fetch = originalFetch;
  }

  assert.equal(calls.length, 1);
  assert.equal(calls[0].url, "https://proxy.example/v1/responses");
  assert.equal(calls[0].body.stream_options, undefined);
  assert.deepEqual(calls[0].body.tools, [{ type: "tool_search" }]);
});

test("DefaultExecutor.transformRequest respects disableStreamOptions for OpenAI chat targets", () => {
  const openAICompat = new DefaultExecutor("openai-compatible-test");
  const chatBody = { model: "gpt-4.1", messages: [{ role: "user", content: "hi" }] };

  const chatResultDisabled = openAICompat.transformRequest("gpt-4.1", chatBody, true, {
    providerSpecificData: { baseUrl: "https://proxy.example/v1", disableStreamOptions: true },
  });

  const chatResultEnabled = openAICompat.transformRequest("gpt-4.1", chatBody, true, {
    providerSpecificData: { baseUrl: "https://proxy.example/v1", disableStreamOptions: false },
  });

  assert.equal((chatResultDisabled as any).stream_options, undefined);
  assert.deepEqual((chatResultEnabled as any).stream_options, { include_usage: true });
});

test("DefaultExecutor.transformRequest injects OpenRouter connection preset", () => {
  const executor = new DefaultExecutor("openrouter");
  const body = { model: "openai/gpt-4", messages: [{ role: "user", content: "hi" }] };

  const result = executor.transformRequest("openai/gpt-4", body, true, {
    providerSpecificData: { preset: "  email-copywriter  " },
  });

  assert.equal((result as any).preset, "email-copywriter");
  assert.deepEqual((result as any).stream_options, { include_usage: true });
  assert.equal((body as any).preset, undefined);

  const explicit = executor.transformRequest(
    "openai/gpt-4",
    { ...body, preset: "client-preset" },
    true,
    { providerSpecificData: { preset: "connection-preset" } }
  );

  assert.equal((explicit as any).preset, "client-preset");

  const explicitNull = executor.transformRequest("openai/gpt-4", { ...body, preset: null }, true, {
    providerSpecificData: { preset: "connection-preset" },
  });
  assert.equal((explicitNull as any).preset, null);

  const explicitEmpty = executor.transformRequest("openai/gpt-4", { ...body, preset: "" }, true, {
    providerSpecificData: { preset: "connection-preset" },
  });
  assert.equal((explicitEmpty as any).preset, "");

  const blank = executor.transformRequest("openai/gpt-4", body, true, {
    providerSpecificData: { preset: "   " },
  });

  assert.equal((blank as any).preset, undefined);
});

test("DefaultExecutor.transformRequest strips stream_options from Anthropic-compatible targets", () => {
  const anthropicCompat = new DefaultExecutor("anthropic-compatible-test");
  const anthropicCcCompat = new DefaultExecutor("anthropic-compatible-cc-test");

  const anthropicBody = {
    model: "claude-sonnet-4-6",
    messages: [{ role: "user", content: "hi" }],
    max_tokens: 1,
    stream_options: { include_usage: true },
  };
  const ccBody = {
    model: "claude-sonnet-4-6",
    messages: [{ role: "user", content: "hi" }],
    max_tokens: 1,
  };

  const anthropicResult = anthropicCompat.transformRequest(
    "claude-sonnet-4-6",
    anthropicBody,
    true,
    {}
  );
  const ccResult = anthropicCcCompat.transformRequest("claude-sonnet-4-6", ccBody, true, {});

  assert.notEqual(anthropicResult, anthropicBody);
  assert.equal((anthropicResult as any).stream_options, undefined);
  assert.equal((ccResult as any).stream_options, undefined);
});

// Port of decolua/9router#1343: openai-compatible-* providers (DeepSeek / Ollama /
// local OpenAI-compatible models) often lack native Structured Output, so a
// `json_schema` response_format is downgraded to `json_object` with the schema
// injected into the system prompt instead.
test("DefaultExecutor.transformRequest downgrades json_schema to json_object for openai-compatible providers and injects the schema into a fresh system prompt", () => {
  const executor = new DefaultExecutor("openai-compatible-deepseek");
  const schema = {
    type: "object",
    properties: { answer: { type: "string" } },
    required: ["answer"],
  };
  const body = {
    model: "deepseek-chat",
    messages: [{ role: "user", content: "give me JSON" }],
    response_format: {
      type: "json_schema",
      json_schema: { name: "answer_schema", schema },
    },
  };

  const result = executor.transformRequest("deepseek-chat", body, true, {
    providerSpecificData: { baseUrl: "https://proxy.example/v1" },
  }) as any;

  // response_format is downgraded to json_object.
  assert.deepEqual(result.response_format, { type: "json_object" });
  // A system message carrying the schema is injected at the front.
  assert.equal(result.messages[0].role, "system");
  assert.match(result.messages[0].content, /strictly follows this JSON schema/);
  assert.ok(result.messages[0].content.includes('"answer"'));
  // The original user message is preserved.
  assert.equal(result.messages[1].role, "user");
  assert.equal(result.messages[1].content, "give me JSON");
  // Original body is not mutated.
  assert.equal((body as any).response_format.type, "json_schema");
  assert.equal(body.messages.length, 1);
});

test("DefaultExecutor.transformRequest appends the json_schema prompt to an existing system message", () => {
  const executor = new DefaultExecutor("openai-compatible-ollama");
  const body = {
    model: "llama3.1",
    messages: [
      { role: "system", content: "You are concise." },
      { role: "user", content: "give me JSON" },
    ],
    response_format: {
      type: "json_schema",
      json_schema: { name: "s", schema: { type: "object" } },
    },
  };

  const result = executor.transformRequest("llama3.1", body, true, {
    providerSpecificData: { baseUrl: "https://proxy.example/v1" },
  }) as any;

  assert.deepEqual(result.response_format, { type: "json_object" });
  assert.equal(result.messages[0].role, "system");
  assert.match(result.messages[0].content, /^You are concise\./);
  assert.match(result.messages[0].content, /strictly follows this JSON schema/);
  // Existing system message object is not mutated in place.
  assert.equal(body.messages[0].content, "You are concise.");
});

// kilocode's DeepSeek V4 Flash rejects ANY `response_format` with HTTP 400
// (verified live 2026-08-15: both json_schema AND json_object 400 with
// `param: response_format`) — same class as the opencode #9992 fix, but the
// default executor's gate only covered `openai-compatible-*`, so kilocode
// forwarded the unsupported format raw. For kilocode the format must be
// STRIPPED entirely (schema injected into the system prompt), because even
// the json_object downgrade is rejected.
test("DefaultExecutor.transformRequest strips response_format for kilocode (DeepSeek 400 regression)", () => {
  const executor = new DefaultExecutor("kilocode");
  const schema = {
    type: "object",
    properties: { answer: { type: "string" } },
    required: ["answer"],
  };
  const body = {
    model: "deepseek/deepseek-v4-flash",
    messages: [{ role: "user", content: "give me JSON" }],
    response_format: {
      type: "json_schema",
      json_schema: { name: "answer_schema", schema },
    },
  };

  const result = executor.transformRequest("deepseek/deepseek-v4-flash", body, true, {
    providerSpecificData: { baseUrl: "https://api.kilo.ai/v1" },
  }) as unknown as {
    response_format?: { type?: string };
    messages: Array<{ role: string; content: string }>;
  };

  // response_format is REMOVED entirely (kilocode rejects json_object too).
  assert.equal(result.response_format, undefined);
  assert.equal(result.messages[0].role, "system");
  assert.match(result.messages[0].content, /strictly follows this JSON schema/);
  assert.ok(result.messages[0].content.includes('"answer"'));
  assert.equal(result.messages[1].role, "user");
  assert.equal(result.messages[1].content, "give me JSON");
  // Original body is not mutated.
  assert.equal(body.response_format.type, "json_schema");
  assert.equal(body.messages.length, 1);
});

test("DefaultExecutor.transformRequest strips response_format for kilocode json_object requests too", () => {
  const executor = new DefaultExecutor("kilocode");
  const body = {
    model: "deepseek/deepseek-v4-flash",
    messages: [{ role: "user", content: "give me JSON" }],
    response_format: { type: "json_object" },
  };

  const result = executor.transformRequest("deepseek/deepseek-v4-flash", body, true, {
    providerSpecificData: { baseUrl: "https://api.kilo.ai/v1" },
  }) as unknown as {
    response_format?: { type?: string };
    messages: Array<{ role: string; content: string }>;
  };

  assert.equal(result.response_format, undefined);
  assert.equal(result.messages[0].role, "system");
  assert.match(result.messages[0].content, /valid JSON only/);
});

test("DefaultExecutor.transformRequest leaves json_schema response_format untouched for native providers", () => {
  const executor = new DefaultExecutor("openai");
  const responseFormat = {
    type: "json_schema",
    json_schema: { name: "s", schema: { type: "object" } },
  };
  const body = {
    model: "gpt-4.1",
    messages: [{ role: "user", content: "give me JSON" }],
    response_format: responseFormat,
  };

  const result = executor.transformRequest("gpt-4.1", body, true, {}) as any;

  // Native OpenAI keeps the json_schema response_format; no system prompt injected.
  assert.deepEqual(result.response_format, responseFormat);
  assert.equal(result.messages.length, 1);
  assert.equal(result.messages[0].role, "user");
});

test("DefaultExecutor.transformRequest ignores non-json_schema response_format for openai-compatible providers", () => {
  const executor = new DefaultExecutor("openai-compatible-deepseek");
  const body = {
    model: "deepseek-chat",
    messages: [{ role: "user", content: "hi" }],
    response_format: { type: "json_object" },
  };

  const result = executor.transformRequest("deepseek-chat", body, true, {
    providerSpecificData: { baseUrl: "https://proxy.example/v1" },
  }) as any;

  assert.deepEqual(result.response_format, { type: "json_object" });
  assert.equal(result.messages.length, 1);
  assert.equal(result.messages[0].role, "user");
});

test("DefaultExecutor.transformRequest applies GLMT preset defaults without overriding explicit values", () => {
  const executor = new DefaultExecutor("glmt");

  const autoBody = {
    messages: [{ role: "user", content: "hi" }],
  };
  const autoResult = executor.transformRequest("glm-5.1", autoBody, true, {});

  assert.notEqual(autoResult, autoBody);
  assert.equal((autoResult as any).max_tokens, 65536);
  (assert as any).equal((autoResult as any).temperature, 0.2);
  (assert as any).deepEqual((autoResult as any).thinking, {
    type: "enabled",
    budget_tokens: 24576,
  });

  const explicitBody = {
    messages: [{ role: "user", content: "hi" }],
    max_tokens: 4096,
    temperature: 0.7,
    thinking: { type: "enabled" },
  };
  const explicitResult = executor.transformRequest("glm-5.1", explicitBody, true, {});

  assert.notEqual(explicitResult, explicitBody);
  assert.equal((explicitResult as any).max_tokens, 4096);
  assert.equal((explicitResult as any).temperature, 0.7);
  assert.deepEqual((explicitResult as any).thinking, {
    type: "enabled",
    budget_tokens: 4095,
  });
});

test("BaseExecutor helpers manage custom user agents and upstream extra headers", () => {
  const headers = { "user-agent": "old", Authorization: "Bearer old" };

  assert.equal(getCustomUserAgent({ customUserAgent: "  MyAgent/1.0  " }), "MyAgent/1.0");
  assert.equal(getCustomUserAgent({ customUserAgent: "   " }), null);

  setUserAgentHeader(headers, "MyAgent/2.0");
  assert.equal(headers["User-Agent"], "MyAgent/2.0");
  assert.equal(headers["user-agent"], undefined);

  applyConfiguredUserAgent(headers, { customUserAgent: "MyAgent/3.0" });
  assert.equal(headers["User-Agent"], "MyAgent/3.0");

  mergeUpstreamExtraHeaders(headers, {
    Authorization: "Bearer override",
    "user-agent": "Merged/4.0",
    "X-Upstream": "1",
  });
  assert.equal(headers.Authorization, "Bearer override");
  assert.equal(headers["User-Agent"], "Merged/4.0");
  assert.equal(headers["user-agent"], undefined);
  assert.equal(headers["X-Upstream"], "1");
});

test("BaseExecutor.mergeAbortSignals aborts when either source signal aborts", () => {
  const primary = new AbortController();
  const secondary = new AbortController();
  const merged = mergeAbortSignals(primary.signal, secondary.signal);

  assert.equal(merged.aborted, false);
  const primaryReason = new Error("primary timeout");
  primaryReason.name = "TimeoutError";
  primary.abort(primaryReason);
  assert.equal(merged.aborted, true);
  assert.equal(merged.reason, primaryReason);

  const otherPrimary = new AbortController();
  const otherSecondary = new AbortController();
  const merged2 = mergeAbortSignals(otherPrimary.signal, otherSecondary.signal);
  const secondaryReason = new Error("client closed");
  otherSecondary.abort(secondaryReason);
  assert.equal(merged2.aborted, true);
  assert.equal(merged2.reason, secondaryReason);
});

test("BaseExecutor.needsRefresh returns true only when expiry is near", () => {
  const executor = new TestExecutor();
  const soon = new Date(Date.now() + 60_000).toISOString();
  const later = new Date(Date.now() + 60 * 60 * 1000).toISOString();

  assert.equal(executor.needsRefresh({ expiresAt: soon }), true);
  assert.equal(executor.needsRefresh({ expiresAt: later }), false);
  assert.equal(executor.needsRefresh({}), false);
});

test("DefaultExecutor.refreshCredentials signals nothing to refresh without refresh token", async () => {
  const executor = new DefaultExecutor("gemini");
  const result = await executor.refreshCredentials({}, null);
  assert.equal(result, NO_REFRESH_NEEDED);
});

test("NO_REFRESH_NEEDED is a frozen shared sentinel", () => {
  assert.equal(Object.isFrozen(NO_REFRESH_NEEDED), true);
  assert.throws(() => {
    "use strict";
    (NO_REFRESH_NEEDED as { error: string }).error = "invalid_grant";
  }, TypeError);
  assert.equal(NO_REFRESH_NEEDED.error, "no_refresh_needed");
});

test("proactive refresh of an API-key connection logs no unrecoverable-sentinel warning", async () => {
  const originalFetch = globalThis.fetch;
  const warnings: string[] = [];
  const sentAuth: string[] = [];
  globalThis.fetch = async (_url, init = {}) => {
    sentAuth.push(String(new Headers(init.headers).get("Authorization") ?? ""));
    return new Response(JSON.stringify({ ok: true }), {
      status: 200,
      headers: { "Content-Type": "application/json" },
    });
  };
  try {
    const executor = new DefaultExecutor("openai-compatible-test");
    // Near expiry → needsRefresh() is true, but there is no refresh token.
    const credentials = {
      apiKey: "test-key",
      expiresAt: new Date(Date.now() + 1_000).toISOString(),
      providerSpecificData: { baseUrl: "https://proxy.example/v1/" },
    };
    assert.equal(executor.needsRefresh(credentials), true);
    await executor.execute({
      model: "gpt-4.1",
      body: { model: "gpt-4.1", messages: [{ role: "user", content: "hi" }] },
      stream: false,
      credentials,
      log: {
        debug: () => {},
        info: () => {},
        warn: (_tag: string, msg: string) => warnings.push(String(msg)),
        error: () => {},
      },
    } as unknown as Parameters<DefaultExecutor["execute"]>[0]);
  } finally {
    globalThis.fetch = originalFetch;
  }
  assert.deepEqual(
    warnings.filter((w) => w.includes("unrecoverable sentinel")),
    [],
    "an API-key connection has nothing to refresh — no sentinel warning"
  );
  assert.equal(sentAuth.length, 1);
  assert.equal(sentAuth[0], "Bearer test-key");
});

test("DefaultExecutor.needsRefresh requests a proactive token for GigaChat", () => {
  const executor = new DefaultExecutor("gigachat");

  assert.equal(executor.needsRefresh({ apiKey: "base64-basic-credentials" }), true);
  assert.equal(
    executor.needsRefresh({
      apiKey: "base64-basic-credentials",
      accessToken: "existing-token",
      expiresAt: new Date(Date.now() + 60 * 60 * 1000).toISOString(),
    }),
    false
  );
});

test("DefaultExecutor.refreshCredentials delegates to OAuth refresh and returns new tokens", async () => {
  const executor = new DefaultExecutor("gemini");
  const originalFetch = globalThis.fetch;
  globalThis.fetch = async (url, options) => {
    assert.match(String(url), /oauth2\.googleapis\.com/);
    assert.equal(options.method, "POST");
    return new Response(
      JSON.stringify({
        access_token: "new-access-token",
        refresh_token: "new-refresh-token",
        expires_in: 3600,
      }),
      {
        status: 200,
        headers: { "Content-Type": "application/json" },
      }
    );
  };

  try {
    const result = await executor.refreshCredentials({ refreshToken: "refresh-me" }, null);
    assert.deepEqual(result, {
      accessToken: "new-access-token",
      refreshToken: "new-refresh-token",
      expiresIn: 3600,
    });
  } finally {
    globalThis.fetch = originalFetch;
  }
});

test("DefaultExecutor.refreshCredentials swallows refresh errors and logs them", async () => {
  const executor = new DefaultExecutor("gemini");
  const originalFetch = globalThis.fetch;
  const messages = [];
  globalThis.fetch = async () => {
    throw new Error("network down");
  };

  try {
    const result = await executor.refreshCredentials(
      { refreshToken: "refresh-me-unrotated" },
      { error: (tag, message) => messages.push({ tag, message }) }
    );
    assert.equal(result, null);
    assert.equal(messages.length, 1);
    assert.match(messages[0].message, /refresh error: network down/);
  } finally {
    globalThis.fetch = originalFetch;
  }
});

test("BaseExecutor.execute returns response metadata and merges headers", async () => {
  const executor = new TestExecutor();
  const originalFetch = globalThis.fetch;
  let captured;
  globalThis.fetch = async (url, options) => {
    captured = { url, options };
    return new Response(JSON.stringify({ ok: true }), {
      status: 200,
      headers: { "Content-Type": "application/json" },
    });
  };

  try {
    const result = await executor.execute({
      model: "gpt-4.1",
      body: { messages: [{ role: "user", content: "hi" }] },
      stream: true,
      credentials: {
        apiKey: "base-key",
        providerSpecificData: { customUserAgent: "CredsAgent/1.0" },
      },
      upstreamExtraHeaders: {
        Authorization: "Bearer override",
        "user-agent": "UpstreamAgent/2.0",
        "X-Trace-Id": "trace-1",
      },
    });

    assert.equal(result.url, "https://primary.example/v1/chat/completions");
    assert.equal(result.response.status, 200);
    (assert as any).equal((result.transformedBody as any).transformed, true);
    assert.equal((result.transformedBody as any).model, "gpt-4.1");
    assert.equal(result.headers.Authorization, "Bearer override");
    assert.equal(result.headers["User-Agent"], "UpstreamAgent/2.0");
    assert.equal(result.headers["user-agent"], undefined);
    assert.equal(result.headers["X-Trace-Id"], "trace-1");
    assert.equal(result.headers.Accept, "text/event-stream");
    assert.equal(captured.options.body.includes('"transformed":true'), true);
  } finally {
    globalThis.fetch = originalFetch;
  }
});

test("BaseExecutor.execute refreshes credentials before the request when needed", async () => {
  class RefreshingExecutor extends BaseExecutor {
    constructor() {
      super("refreshing-provider", {
        baseUrl: "https://refresh.example/v1/chat/completions",
      });
    }

    needsRefresh() {
      return true;
    }

    async refreshCredentials() {
      return {
        accessToken: "fresh-token",
        expiresAt: new Date(Date.now() + 60 * 60 * 1000).toISOString(),
      };
    }
  }

  const executor = new RefreshingExecutor();
  const originalFetch = globalThis.fetch;
  let capturedHeaders;
  globalThis.fetch = async (url, options) => {
    assert.equal(String(url), "https://refresh.example/v1/chat/completions");
    capturedHeaders = options.headers;
    return new Response(JSON.stringify({ ok: true }), { status: 200 });
  };

  try {
    await executor.execute({
      model: "gpt-4.1",
      body: {},
      stream: false,
      credentials: { apiKey: "stale-token" },
    });

    assert.equal(capturedHeaders.Authorization, "Bearer fresh-token");
  } finally {
    globalThis.fetch = originalFetch;
  }
});

test("BaseExecutor.execute falls back to the next base URL after a transport error", async () => {
  const executor = new TestExecutor();
  const originalFetch = globalThis.fetch;
  const calls = [];
  globalThis.fetch = async (url) => {
    calls.push(String(url));
    if (calls.length === 1) {
      throw new Error("first node down");
    }
    return new Response("ok", { status: 200 });
  };

  try {
    const result = await executor.execute({
      model: "gpt-4.1",
      body: { hello: "world" },
      stream: false,
      credentials: {},
    });

    assert.deepEqual(calls, [
      "https://primary.example/v1/chat/completions",
      "https://fallback.example/v1/chat/completions",
    ]);
    assert.equal(result.url, "https://fallback.example/v1/chat/completions");
  } finally {
    globalThis.fetch = originalFetch;
  }
});

test("BaseExecutor.execute throws the last error when all URLs fail", async () => {
  const executor = new TestExecutor();
  const originalFetch = globalThis.fetch;
  globalThis.fetch = async () => {
    throw new Error("still down");
  };

  try {
    await assert.rejects(
      executor.execute({
        model: "gpt-4.1",
        body: {},
        stream: false,
        credentials: {},
      }),
      /still down/
    );
  } finally {
    globalThis.fetch = originalFetch;
  }
});

test("BaseExecutor.execute propagates aborted requests through the merged signal", async () => {
  const executor = new TestExecutor({ baseUrls: ["https://single.example/v1/chat/completions"] });
  const controller = new AbortController();
  controller.abort();
  const originalFetch = globalThis.fetch;

  globalThis.fetch = async (url, options) => {
    assert.equal(options.signal.aborted, true);
    const error = new Error(`aborted ${url}`);
    error.name = "AbortError";
    throw error;
  };

  try {
    await assert.rejects(
      executor.execute({
        model: "gpt-4.1",
        body: {},
        stream: false,
        credentials: {},
        signal: controller.signal,
      }),
      /aborted/
    );
  } finally {
    globalThis.fetch = originalFetch;
  }
});

test("BaseExecutor.execute clears the startup timeout after headers arrive", async () => {
  const executor = new TestExecutor({ baseUrls: ["https://single.example/v1/chat/completions"] });
  const originalFetch = globalThis.fetch;
  const originalFetchStartTimeoutMs = BaseExecutor.FETCH_START_TIMEOUT_MS;
  let capturedSignal;

  BaseExecutor.FETCH_START_TIMEOUT_MS = 20;
  globalThis.fetch = async (_url, options) => {
    capturedSignal = options.signal;
    return new Response("ok", {
      status: 200,
      headers: { "Content-Type": "application/json" },
    });
  };

  try {
    await executor.execute({
      model: "gpt-4.1",
      body: {},
      stream: true,
      credentials: {},
    });

    assert.equal(capturedSignal?.aborted, false);
    await new Promise((resolve) => setTimeout(resolve, 40));
    assert.equal(capturedSignal?.aborted, false);
  } finally {
    BaseExecutor.FETCH_START_TIMEOUT_MS = originalFetchStartTimeoutMs;
    globalThis.fetch = originalFetch;
  }
});

// Regression test for issue #1454: duplicate anthropic-version header when
// Claude Code CLI headers are detected on the native `claude` provider.
// The provider config seeds headers with Title-Case "Anthropic-Version" while
// the Claude-Code patch injects lowercase "anthropic-version".  Before the fix,
// both keys coexisted in the JS object and undici combined their values into
// "2023-06-01, 2023-06-01", causing a 400 from Anthropic.
test("DefaultExecutor.execute does not produce duplicate anthropic-version header when Claude Code CLI headers are present", async () => {
  const executor = new DefaultExecutor("claude");
  const originalFetch = globalThis.fetch;
  let capturedHeaders: Record<string, string> = {};
  let capturedBody = "";

  globalThis.fetch = async (_url, init = {}) => {
    // Capture raw headers without normalisation so case-variant duplicate keys are visible.
    capturedHeaders = (init.headers as Record<string, string>) || {};
    capturedBody = String(init.body ?? "");
    return new Response(JSON.stringify({ ok: true }), {
      status: 200,
      headers: { "Content-Type": "application/json" },
    });
  };

  try {
    await executor.execute({
      model: "claude-sonnet-4-6",
      body: {
        model: "claude-sonnet-4-6",
        messages: [{ role: "user", content: "hi" }],
        max_tokens: 1,
      },
      stream: false,
      credentials: { accessToken: "oauth-token" },
      clientHeaders: {
        "x-app": "cli",
        "user-agent": "claude-cli/2.1.116 (external, cli)",
        "anthropic-beta": "oauth-2025-04-20",
      },
    });
  } finally {
    globalThis.fetch = originalFetch;
  }

  // Must be exactly one key — not multiple case variants that undici would combine
  const versionKeys = Object.keys(capturedHeaders).filter(
    (k) => k.toLowerCase() === "anthropic-version"
  );
  assert.equal(versionKeys.length, 1, "Duplicate anthropic-version header keys found");
  assert.equal(capturedHeaders[versionKeys[0]], "2023-06-01");
  assert.equal(capturedHeaders["X-Stainless-Runtime-Version"], "v26.3.0");
  assert.equal(capturedHeaders["X-Stainless-Package-Version"], "0.112.1");

  const sentBody = JSON.parse(capturedBody) as { system?: Array<{ text?: string }> };
  const cc = `x-anthropic-billing-header: cc_version=${CLAUDE_CODE_CLIENT_BILLING_VERSION}; `;
  const billing = sentBody.system?.[0]?.text ?? "";
  assert.equal(billing.slice(0, cc.length), cc, "cc_version tracks the constant (#14627)");
  assert.match(billing.slice(cc.length), /^cc_entrypoint=cli; cch=[0-9a-f]{5};$/);
});

test('shouldForceResponsesUpstream respects explicit apiType="chat" even when namespace tools are present', () => {
  const body = {
    input: "hi",
    tools: [
      {
        type: "namespace",
        name: "collaboration",
        tools: [
          {
            name: "spawn_agent",
            description: "Spawn an agent",
            parameters: { type: "object", properties: { task: { type: "string" } } },
          },
        ],
      },
    ],
  };
  const credentials = {
    providerSpecificData: {
      baseUrl: "https://ark.cn-beijing.volces.com/api/coding/v3",
      apiType: "chat",
    },
  };
  assert.equal(
    shouldForceResponsesUpstream("openai-compatible-responses-demo", body, credentials),
    false
  );
});

test("shouldForceResponsesUpstream still forces /responses for untyped OpenAI-compatible providers with namespace tools", () => {
  const body = {
    input: "hi",
    tools: [
      {
        type: "namespace",
        name: "collaboration",
        tools: [
          { name: "spawn_agent", description: "Spawn an agent", parameters: { type: "object" } },
        ],
      },
    ],
  };
  const credentials = {
    providerSpecificData: {
      baseUrl: "https://proxy.example/v1",
    },
  };
  assert.equal(shouldForceResponsesUpstream("openai-compatible-test", body, credentials), true);
});