@jeffreycao/copilot-api 1.10.7 → 1.10.9

package/README.md

@@ -363,7 +363,6 @@ The following command line options are available for the `start` command:
       "gpt-5.4-mini": "xhigh",
       "gpt-5.4": "xhigh"
     },
-    "useFunctionApplyPatch": true,
     "useMessagesApi": true,
     "useResponsesApiWebSocket": true,
     "useResponsesApiWebSearch": true
@@ -383,14 +382,13 @@ The following command line options are available for the `start` command:
     - `temperature` (optional): Default temperature value used when the request does not specify one.
     - `topP` (optional): Default top_p value used when the request does not specify one.
     - `topK` (optional): Default top_k value used when the request does not specify one.
-    - `extraBody` (optional): Dynamic fields merged into the upstream request body for that model. Request body fields with the same name take precedence. OpenAI-compatible providers can use this for fields such as `enable_thinking`, `preserve_thinking`, `reasoning_effort`.
+    - `extraBody` (optional): Dynamic fields merged into the upstream request body for that model. Request body fields with the same name take precedence. OpenAI-compatible providers can use this for fields such as `enable_thinking`, `preserve_thinking`, `reasoning_effort`. `thinking_budget` is a special OpenAI-compatible provider override: when configured in `extraBody`, it is forced after Anthropic `thinking.budget_tokens` translation and overrides the request-derived budget.
     - `contextCache` (optional): Defaults to `true` for OpenAI-compatible providers. This enables Alibaba Cloud Model Studio/DashScope explicit context cache by injecting `cache_control: { "type": "ephemeral" }` on up to 4 content blocks using the Context Cache format. The cache breakpoint strategy matches opencode's main provider flow: the first 2 system messages plus the last 2 non-system messages. Marked string content is converted to text content part arrays for `system` / `user` / `assistant` / `tool` messages; existing array content is marked on the last part. Set this to `false` when the model already supports implicit caching, or when the upstream does not accept this explicit-cache extension field.
     - `supportPdf` (optional): Controls whether the model supports PDF/document content. Defaults to `false`; unsupported PDFs are converted to a text notice. Set it to `true` to send PDF/document blocks as OpenAI Chat Completions file parts.
     - `toolContentSupportType` (optional): Tool result content capabilities for that model, as an array of `array`, `image`, and `pdf`. Provider routes default to string-only tool content when omitted. If `supportPdf` is `true` but this list does not include `pdf`, file parts in tool results are moved to user role messages. This provider default does not change the Copilot main flow, which continues to support array + image and not PDF.
 - **smallModel:** Fallback model used for tool-less warmup messages (e.g., Claude Code probe requests) to avoid spending premium requests; defaults to gpt-5-mini.
 - **responsesApiContextManagementModels:** List of GPT model IDs that should receive Responses API `context_management` compaction instructions. This defaults to `[]`, so you need to opt in explicitly. A good starting point is `["gpt-5-mini", "gpt-5.3-codex", "gpt-5.4-mini", "gpt-5.4"]`. When enabled, the request includes `context_management` in the body and keeps only the latest compaction carrier on follow-up turns. The actual compaction is handled server-side and appears to begin when usage approaches roughly 90% of the model's `maxPromptTokens`, which makes it especially useful for long-running tasks without consuming additional premium requests. In practice, the effective `compact_threshold` also appears to be fixed on the server side, so changing it in this project does not currently alter compaction behavior. At the moment, this optimization is intended for GPT-family models only.
 - **modelReasoningEfforts:** Per-model `reasoning.effort` sent to the Copilot Responses API. Allowed values are `none`, `minimal`, `low`, `medium`, `high`, and `xhigh`. If a model isn’t listed, `high` is used by default.
-- **useFunctionApplyPatch:** When `true`, the server will convert any custom tool named `apply_patch` in Responses payloads into an OpenAI-style function tool (`type: "function"`) with a parameter schema so assistants can call it using function-calling semantics to edit files. Set to `false` to leave tools unchanged. Defaults to `true`.
 - **useMessagesApi:** When `true`, Claude-family models that support Copilot's native `/v1/messages` endpoint will use the Messages API; otherwise they fall back to `/chat/completions`. Set to `false` to disable Messages API routing and always use `/chat/completions`. Defaults to `true`.
 - **useResponsesApiWebSocket:** When `true`, Responses API requests use Copilot's websocket transport for models that advertise `ws:/responses`; models that only advertise `/responses` continue to use HTTP. Set to `false` to disable websocket routing and use HTTP `/responses` whenever the selected model supports it. Defaults to `true`.
 - **useResponsesApiWebSearch:** When `true`, the server keeps Responses API tools with `type: "web_search"` and forwards them upstream. Set to `false` to strip those tools from `/responses` payloads. Defaults to `true`.

package/README.zh-CN.md

@@ -367,7 +367,6 @@ Copilot API 现在使用子命令结构，主要命令包括：
       "gpt-5.4-mini": "xhigh",
       "gpt-5.4": "xhigh"
     },
-    "useFunctionApplyPatch": true,
     "useMessagesApi": true,
     "useResponsesApiWebSocket": true,
     "useResponsesApiWebSearch": true
@@ -387,14 +386,13 @@ Copilot API 现在使用子命令结构，主要命令包括：
     - `temperature`：可选，当请求未指定时使用的默认温度。
     - `topP`：可选，当请求未指定时使用的默认 `top_p`。
     - `topK`：可选，当请求未指定时使用的默认 `top_k`。
-    - `extraBody`：可选，按模型合入上游请求体的动态字段；请求体显式同名字段优先。OpenAI 兼容 provider 可用它配置 `enable_thinking`、`preserve_thinking`、`reasoning_effort` 等字段。
+    - `extraBody`：可选，按模型合入上游请求体的动态字段；请求体显式同名字段优先。OpenAI 兼容 provider 可用它配置 `enable_thinking`、`preserve_thinking`、`reasoning_effort` 等字段。`thinking_budget` 是 OpenAI 兼容 provider 的特殊覆盖项：配置在 `extraBody` 后，会在 Anthropic `thinking.budget_tokens` 翻译之后强制写入，并覆盖请求派生出的预算值。
     - `contextCache`：可选，OpenAI 兼容 provider 默认 `true`，用于启用阿里云百炼/DashScope 的显式缓存（explicit context cache），会按其 Context Cache 格式在最多 4 个 content block 上注入 `cache_control: { "type": "ephemeral" }`。缓存断点策略与 opencode 主链路保持一致：前 2 条 system 消息 + 最后 2 条非 system 消息。标记字符串 content 时会把 `system` / `user` / `assistant` / `tool` 消息转换为 text content part 数组；已有数组 content 则标记最后一个 part。如果模型本身已经支持隐式缓存，或上游不支持该显式缓存扩展字段，可在模型配置中设为 `false`。
     - `supportPdf`：可选，控制该模型是否支持 PDF/document content。默认 `false`，不支持时会把 PDF 转成提示文本；设为 `true` 时会把 PDF/document 转成 OpenAI Chat Completions 的 file part。
     - `toolContentSupportType`：可选，配置该模型的 tool result content 支持能力，值为 `array`、`image`、`pdf` 的数组。provider 侧未配置时默认只发送 string tool content。若 `supportPdf` 为 `true` 但这里不包含 `pdf`，tool result 里的 file part 会被转成 user role 消息。Copilot 主链路不使用这个 provider 默认，仍按 array + image 且不支持 PDF 的能力处理。
 - **smallModel：** 无工具预热消息的回退模型（例如 Claude Code 的探测请求），用于避免消耗 premium requests；默认是 `gpt-5-mini`。
 - **responsesApiContextManagementModels：** 需要启用 Responses API `context_management` 压缩指令的 GPT 模型 ID 列表。默认是 `[]`，需要你显式开启。一个不错的起点是 `["gpt-5-mini", "gpt-5.3-codex", "gpt-5.4-mini", "gpt-5.4"]`。启用后，请求体会带上 `context_management`，并在后续轮次中仅保留最新的压缩承载内容。实际压缩由服务端完成，看起来会在 usage 接近模型 `maxPromptTokens` 的约 90% 时开始，因此特别适合长任务场景，同时不会额外消耗 premium requests。实践中 `compact_threshold` 似乎也是服务端固定的，所以在本项目中修改它目前不会改变压缩行为。当前该优化仅面向 GPT 系模型。
 - **modelReasoningEfforts：** 按模型配置发送到 Copilot Responses API 的 `reasoning.effort`。可选值包括 `none`、`minimal`、`low`、`medium`、`high` 和 `xhigh`。若某模型未配置，则默认使用 `high`。
-- **useFunctionApplyPatch：** 当为 `true` 时，服务端会把 Responses payload 中任何名为 `apply_patch` 的自定义工具转换为 OpenAI 风格的函数工具（`type: "function"`），并附带参数 schema，从而让 assistant 可以通过 function-calling 语义调用它来编辑文件。若设为 `false`，则保持工具原样。默认值为 `true`。
 - **useMessagesApi：** 当为 `true` 时，支持 Copilot 原生 `/v1/messages` 的 Claude 系模型会走 Messages API；否则回退到 `/chat/completions`。设为 `false` 可禁用 Messages API 路由，始终使用 `/chat/completions`。默认值为 `true`。
 - **useResponsesApiWebSocket：** 当为 `true` 时，Responses API 请求会优先对声明了 `ws:/responses` 的模型使用 Copilot websocket transport；仅声明 `/responses` 的模型仍走 HTTP。设为 `false` 可禁用 websocket 路由，并在模型支持 `/responses` 时使用 HTTP `/responses`。默认值为 `true`。
 - **useResponsesApiWebSearch：** 当为 `true` 时，服务端会保留 Responses API 中 `type: "web_search"` 的工具并透传到上游。设为 `false` 则会从 `/responses` payload 中移除这些工具。默认值为 `true`。

package/dist/main.js

@@ -43,7 +43,7 @@ const { auth } = await import("./auth-BHa2OHXf.js");
 const { checkUsage } = await import("./check-usage-BdXGp1Wr.js");
 const { debug } = await import("./debug-C_TBkyUw.js");
 const { mcp } = await import("./mcp-CTb-DbQH.js");
-const { start } = await import("./start-D5EioAT0.js");
+const { start } = await import("./start-CbKg_0bY.js");
 await runMain(defineCommand({
 	meta: {
 		name: "copilot-api",

package/dist/{proxy-CSO5SLia.js → proxy-DvlF9a-7.js}

@@ -50,7 +50,6 @@ const defaultConfig = {
 		"gpt-5.4": "xhigh",
 		"gpt-5.5": "xhigh"
 	},
-	useFunctionApplyPatch: true,
 	useMessagesApi: true,
 	useResponsesApiWebSocket: true,
 	useResponsesApiWebSearch: true
@@ -350,4 +349,4 @@ function initProxyFromEnv() {
 //#endregion
 export { setModelMappings as _, getConfig as a, getProviderConfig as c, isMessagesApiEnabled as d, isResponsesApiContextManagementModel as f, resolveMappedModel as g, mergeConfigWithDefaults as h, getClaudeTokenMultiplier as i, getReasoningEffortForModel as l, isResponsesApiWebSocketEnabled as m, initProxyFromEnv as n, getExtraPromptForModel as o, isResponsesApiWebSearchEnabled as p, getAnthropicApiKey as r, getModelMappings as s, getProxyEnvDispatcher as t, getSmallModel as u };
 
-//# sourceMappingURL=proxy-CSO5SLia.js.map
+//# sourceMappingURL=proxy-DvlF9a-7.js.map

package/dist/proxy-DvlF9a-7.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"proxy-DvlF9a-7.js","names":["get"],"sources":["../src/lib/config.ts","../src/lib/proxy.ts"],"sourcesContent":["import consola from \"consola\"\nimport { randomBytes } from \"node:crypto\"\nimport fs from \"node:fs\"\n\nimport { PATHS } from \"./paths\"\n\nexport interface AppConfig {\n  auth?: {\n    apiKeys?: Array<string>\n    adminApiKey?: string\n  }\n  providers?: Record<string, ProviderConfig>\n  modelMappings?: Record<string, string>\n  extraPrompts?: Record<string, string>\n  smallModel?: string\n  responsesApiContextManagementModels?: Array<string>\n  modelReasoningEfforts?: Record<\n    string,\n    \"none\" | \"minimal\" | \"low\" | \"medium\" | \"high\" | \"xhigh\"\n  >\n  useMessagesApi?: boolean\n  useResponsesApiWebSocket?: boolean\n  anthropicApiKey?: string\n  useResponsesApiWebSearch?: boolean\n  claudeTokenMultiplier?: number\n}\n\nexport interface ModelConfig {\n  temperature?: number\n  topP?: number\n  topK?: number\n  extraBody?: Record<string, unknown>\n  contextCache?: boolean\n  supportPdf?: boolean\n  toolContentSupportType?: Array<ToolContentSupportType>\n}\n\nexport type ProviderAuthType = \"authorization\" | \"x-api-key\"\nexport type ProviderType = \"anthropic\" | \"openai-compatible\"\nexport type ToolContentSupportType = \"array\" | \"image\" | \"pdf\"\n\nexport interface ProviderConfig {\n  type?: string\n  enabled?: boolean\n  baseUrl?: string\n  apiKey?: string\n  authType?: ProviderAuthType\n  models?: Record<string, ModelConfig>\n  adjustInputTokens?: boolean\n}\n\nexport interface ResolvedProviderConfig {\n  name: string\n  type: ProviderType\n  baseUrl: string\n  apiKey: string\n  authType: ProviderAuthType\n  models?: Record<string, ModelConfig>\n  adjustInputTokens?: boolean\n}\n\nconst gpt5ExplorationPrompt = `## Exploration and reading files\n- **Think first.** Before any tool call, decide ALL files/resources you will need.\n- **Batch everything.** If you need multiple files (even from different places), read them together.\n- **multi_tool_use.parallel** Use multi_tool_use.parallel to parallelize tool calls and only this.\n- **Only make sequential calls if you truly cannot know the next file without seeing a result first.**\n- **Workflow:** (a) plan all needed reads → (b) issue one parallel batch → (c) analyze results → (d) repeat if new, unpredictable reads arise.`\n\nconst gpt5CommentaryPrompt = `# Working with the user\n\nYou interact with the user through a terminal. You have 2 ways of communicating with the users:  \n- Share intermediary updates in \\`commentary\\` channel.  \n- After you have completed all your work, send a message to the \\`final\\` channel.  \n\n## Intermediary updates\n\n- Intermediary updates go to the \\`commentary\\` channel.\n- User updates are short updates while you are working, they are NOT final answers.\n- You use 1-2 sentence user updates to communicate progress and new information to the user as you are doing work.\n- Do not begin responses with conversational interjections or meta commentary. Avoid openers such as acknowledgements (“Done —”, “Got it”, “Great question, ”) or framing phrases.\n- You provide user updates frequently, every 20s.\n- Before exploring or doing substantial work, you start with a user update acknowledging the request and explaining your first step. You should include your understanding of the user request and explain what you will do. Avoid commenting on the request or using starters such as \"Got it -\" or \"Understood -\" etc.\n- When exploring, e.g. searching, reading files, you provide user updates as you go, every 20s, explaining what context you are gathering and what you've learned. Vary your sentence structure when providing these updates to avoid sounding repetitive - in particular, don't start each sentence the same way.\n- After you have sufficient context, and the work is substantial, you provide a longer plan (this is the only user update that may be longer than 2 sentences and can contain formatting).\n- Before performing file edits of any kind, you provide updates explaining what edits you are making.\n- As you are thinking, you very frequently provide updates even if not taking any actions, informing the user of your progress. You interrupt your thinking and send multiple updates in a row if thinking for more than 100 words.\n- Tone of your updates MUST match your personality.`\n\nconst defaultConfig: AppConfig = {\n  auth: {\n    apiKeys: [],\n  },\n  providers: {},\n  modelMappings: {},\n  extraPrompts: {\n    \"gpt-5-mini\": gpt5ExplorationPrompt,\n    \"gpt-5.3-codex\": gpt5CommentaryPrompt,\n    \"gpt-5.4-mini\": gpt5CommentaryPrompt,\n    \"gpt-5.4\": gpt5CommentaryPrompt,\n    \"gpt-5.5\": gpt5CommentaryPrompt,\n  },\n  smallModel: \"gpt-5-mini\",\n  responsesApiContextManagementModels: [],\n  modelReasoningEfforts: {\n    \"gpt-5-mini\": \"low\",\n    \"gpt-5.3-codex\": \"xhigh\",\n    \"gpt-5.4-mini\": \"xhigh\",\n    \"gpt-5.4\": \"xhigh\",\n    \"gpt-5.5\": \"xhigh\",\n  },\n  useMessagesApi: true,\n  useResponsesApiWebSocket: true,\n  useResponsesApiWebSearch: true,\n}\n\nlet cachedConfig: AppConfig | null = null\n\nfunction normalizeAdminApiKey(adminApiKey: unknown): string | null {\n  if (typeof adminApiKey !== \"string\") {\n    if (adminApiKey !== undefined) {\n      consola.warn(\n        \"Invalid auth.adminApiKey config. Expected a non-empty string.\",\n      )\n    }\n    return null\n  }\n\n  const normalizedAdminApiKey = adminApiKey.trim()\n  if (!normalizedAdminApiKey) {\n    consola.warn(\n      \"Invalid auth.adminApiKey config. Expected a non-empty string.\",\n    )\n    return null\n  }\n\n  return normalizedAdminApiKey\n}\n\nfunction generateAdminApiKey(): string {\n  return randomBytes(32).toString(\"hex\")\n}\n\nfunction isNodeError(error: unknown): error is NodeJS.ErrnoException {\n  return error instanceof Error && \"code\" in error\n}\n\nfunction ensureConfigFile(): void {\n  try {\n    fs.accessSync(PATHS.CONFIG_PATH, fs.constants.R_OK | fs.constants.W_OK)\n  } catch {\n    fs.mkdirSync(PATHS.APP_DIR, { recursive: true })\n    fs.writeFileSync(\n      PATHS.CONFIG_PATH,\n      `${JSON.stringify(defaultConfig, null, 2)}\\n`,\n      \"utf8\",\n    )\n    try {\n      fs.chmodSync(PATHS.CONFIG_PATH, 0o600)\n    } catch {\n      return\n    }\n  }\n}\n\nfunction readConfigFromDisk(): AppConfig {\n  ensureConfigFile()\n  try {\n    const raw = fs.readFileSync(PATHS.CONFIG_PATH, \"utf8\")\n    if (!raw.trim()) {\n      fs.writeFileSync(\n        PATHS.CONFIG_PATH,\n        `${JSON.stringify(defaultConfig, null, 2)}\\n`,\n        \"utf8\",\n      )\n      return defaultConfig\n    }\n    return JSON.parse(raw) as AppConfig\n  } catch (error) {\n    consola.error(\"Failed to read config file, using default config\", error)\n    return defaultConfig\n  }\n}\n\nfunction readEditableConfigFromDisk(): AppConfig {\n  try {\n    const raw = fs.readFileSync(PATHS.CONFIG_PATH, \"utf8\")\n    if (!raw.trim()) {\n      return {}\n    }\n    return JSON.parse(raw) as AppConfig\n  } catch (error) {\n    if (isNodeError(error) && error.code === \"ENOENT\") {\n      return {}\n    }\n    if (error instanceof SyntaxError) {\n      throw new Error(`Config file is not valid JSON: ${PATHS.CONFIG_PATH}`)\n    }\n    throw error\n  }\n}\n\nfunction writeConfigToDisk(config: AppConfig): void {\n  fs.mkdirSync(PATHS.APP_DIR, { recursive: true })\n  fs.writeFileSync(\n    PATHS.CONFIG_PATH,\n    `${JSON.stringify(config, null, 2)}\\n`,\n    \"utf8\",\n  )\n}\n\nfunction mergeDefaultConfig(config: AppConfig): {\n  mergedConfig: AppConfig\n  changed: boolean\n} {\n  const extraPrompts = config.extraPrompts ?? {}\n  const defaultExtraPrompts = defaultConfig.extraPrompts ?? {}\n  const modelReasoningEfforts = config.modelReasoningEfforts ?? {}\n  const defaultModelReasoningEfforts = defaultConfig.modelReasoningEfforts ?? {}\n\n  const missingExtraPromptModels = Object.keys(defaultExtraPrompts).filter(\n    (model) => !Object.hasOwn(extraPrompts, model),\n  )\n\n  const missingReasoningEffortModels = Object.keys(\n    defaultModelReasoningEfforts,\n  ).filter((model) => !Object.hasOwn(modelReasoningEfforts, model))\n\n  const hasExtraPromptChanges = missingExtraPromptModels.length > 0\n  const hasReasoningEffortChanges = missingReasoningEffortModels.length > 0\n\n  if (!hasExtraPromptChanges && !hasReasoningEffortChanges) {\n    return { mergedConfig: config, changed: false }\n  }\n\n  return {\n    mergedConfig: {\n      ...config,\n      extraPrompts: {\n        ...defaultExtraPrompts,\n        ...extraPrompts,\n      },\n      modelReasoningEfforts: {\n        ...defaultModelReasoningEfforts,\n        ...modelReasoningEfforts,\n      },\n    },\n    changed: true,\n  }\n}\n\nfunction ensureAdminApiKey(config: AppConfig): {\n  mergedConfig: AppConfig\n  changed: boolean\n} {\n  const normalizedAdminApiKey = normalizeAdminApiKey(config.auth?.adminApiKey)\n  if (normalizedAdminApiKey) {\n    if (config.auth?.adminApiKey === normalizedAdminApiKey) {\n      return { mergedConfig: config, changed: false }\n    }\n\n    return {\n      mergedConfig: {\n        ...config,\n        auth: {\n          ...(config.auth ?? {}),\n          adminApiKey: normalizedAdminApiKey,\n        },\n      },\n      changed: true,\n    }\n  }\n\n  const editableConfig = readEditableConfigFromDisk()\n  const { mergedConfig } = mergeDefaultConfig({\n    ...editableConfig,\n    auth: {\n      ...(editableConfig.auth ?? {}),\n      adminApiKey: generateAdminApiKey(),\n    },\n  })\n\n  return { mergedConfig, changed: true }\n}\n\nexport function mergeConfigWithDefaults(): AppConfig {\n  const config = readConfigFromDisk()\n  const { mergedConfig, changed } = mergeDefaultConfig(config)\n  const {\n    mergedConfig: mergedConfigWithAdminApiKey,\n    changed: adminApiKeyChanged,\n  } = ensureAdminApiKey(mergedConfig)\n  const shouldPersistConfig = changed || adminApiKeyChanged\n\n  if (shouldPersistConfig) {\n    try {\n      writeConfigToDisk(mergedConfigWithAdminApiKey)\n    } catch (writeError) {\n      if (adminApiKeyChanged) {\n        throw writeError\n      }\n\n      consola.warn(\n        \"Failed to write merged extraPrompts to config file\",\n        writeError,\n      )\n    }\n  }\n\n  cachedConfig = mergedConfigWithAdminApiKey\n  return mergedConfigWithAdminApiKey\n}\n\nexport function getConfig(): AppConfig {\n  cachedConfig ??= mergeDefaultConfig(readConfigFromDisk()).mergedConfig\n  return cachedConfig\n}\n\nexport function reloadConfig(): AppConfig {\n  return mergeConfigWithDefaults()\n}\n\nexport function getExtraPromptForModel(model: string): string {\n  const config = getConfig()\n  return config.extraPrompts?.[model] ?? \"\"\n}\n\nexport function getModelMappings(): Record<string, string> {\n  const config = getConfig()\n  const modelMappings = config.modelMappings\n  if (!modelMappings) {\n    return { ...(defaultConfig.modelMappings ?? {}) }\n  }\n\n  const validMappings: Record<string, string> = {}\n  for (const [sourceModel, targetModel] of Object.entries(modelMappings)) {\n    if (\n      !sourceModel\n      || typeof targetModel !== \"string\"\n      || targetModel.length === 0\n    ) {\n      continue\n    }\n    validMappings[sourceModel] = targetModel\n  }\n\n  return validMappings\n}\n\nfunction validateModelMappings(\n  modelMappings: Record<string, string>,\n): Record<string, string> {\n  const validatedMappings: Record<string, string> = {}\n  for (const [sourceModel, targetModel] of Object.entries(modelMappings)) {\n    if (!sourceModel || !targetModel) {\n      throw new Error(\n        \"Each model mapping must use non-empty source and target values.\",\n      )\n    }\n    validatedMappings[sourceModel] = targetModel\n  }\n\n  return validatedMappings\n}\n\nexport function setModelMappings(\n  modelMappings: Record<string, string>,\n): Record<string, string> {\n  const nextConfig = {\n    ...readEditableConfigFromDisk(),\n    modelMappings: validateModelMappings(modelMappings),\n  }\n\n  writeConfigToDisk(nextConfig)\n  cachedConfig = reloadConfig()\n  return getModelMappings()\n}\n\nexport function resolveMappedModel(model: string): string {\n  return getModelMappings()[model] ?? model\n}\n\nexport function getSmallModel(): string {\n  const config = getConfig()\n  return config.smallModel ?? \"gpt-5-mini\"\n}\n\nexport function getResponsesApiContextManagementModels(): Array<string> {\n  const config = getConfig()\n  return (\n    config.responsesApiContextManagementModels\n    ?? defaultConfig.responsesApiContextManagementModels\n    ?? []\n  )\n}\n\nexport function isResponsesApiContextManagementModel(model: string): boolean {\n  return getResponsesApiContextManagementModels().includes(model)\n}\n\nexport function getReasoningEffortForModel(\n  model: string,\n): \"none\" | \"minimal\" | \"low\" | \"medium\" | \"high\" | \"xhigh\" {\n  const config = getConfig()\n  return config.modelReasoningEfforts?.[model] ?? \"high\"\n}\n\nexport function normalizeProviderBaseUrl(url: string): string {\n  return url.trim().replace(/\\/+$/u, \"\")\n}\n\nfunction getDefaultProviderAuthType(\n  providerType: ProviderType,\n): ProviderAuthType {\n  return providerType === \"openai-compatible\" ? \"authorization\" : \"x-api-key\"\n}\n\nexport function resolveProviderAuthType(\n  providerName: string,\n  authType: string | undefined,\n  providerType: ProviderType,\n): ProviderAuthType {\n  if (authType === undefined) {\n    return getDefaultProviderAuthType(providerType)\n  }\n\n  if (authType === \"x-api-key\") {\n    return \"x-api-key\"\n  }\n\n  if (authType === \"authorization\") {\n    return authType\n  }\n\n  consola.warn(\n    `Provider ${providerName} has invalid authType '${authType}', falling back to ${getDefaultProviderAuthType(providerType)}`,\n  )\n  return getDefaultProviderAuthType(providerType)\n}\n\nexport function getProviderConfig(name: string): ResolvedProviderConfig | null {\n  const providerName = name.trim()\n  if (!providerName) {\n    return null\n  }\n\n  const config = getConfig()\n  const provider = config.providers?.[providerName]\n  if (!provider) {\n    return null\n  }\n\n  if (provider.enabled === false) {\n    return null\n  }\n\n  const type = provider.type ?? \"anthropic\"\n  if (type !== \"anthropic\" && type !== \"openai-compatible\") {\n    consola.warn(\n      `Provider ${providerName} is ignored because type '${type}' is not supported`,\n    )\n    return null\n  }\n\n  const baseUrl = normalizeProviderBaseUrl(provider.baseUrl ?? \"\")\n  const apiKey = (provider.apiKey ?? \"\").trim()\n  const authType = resolveProviderAuthType(\n    providerName,\n    provider.authType,\n    type,\n  )\n  if (!baseUrl || !apiKey) {\n    consola.warn(\n      `Provider ${providerName} is enabled but missing baseUrl or apiKey`,\n    )\n    return null\n  }\n\n  return {\n    name: providerName,\n    type,\n    baseUrl,\n    apiKey,\n    authType,\n    models: provider.models,\n    adjustInputTokens: provider.adjustInputTokens,\n  }\n}\n\nexport function listEnabledProviders(): Array<string> {\n  const config = getConfig()\n  const providerNames = Object.keys(config.providers ?? {})\n  return providerNames.filter((name) => getProviderConfig(name) !== null)\n}\n\nexport function isMessagesApiEnabled(): boolean {\n  const config = getConfig()\n  return config.useMessagesApi ?? true\n}\n\nexport function isResponsesApiWebSocketEnabled(): boolean {\n  const config = getConfig()\n  return config.useResponsesApiWebSocket ?? true\n}\n\nexport function getAnthropicApiKey(): string | undefined {\n  const config = getConfig()\n  return config.anthropicApiKey ?? process.env.ANTHROPIC_API_KEY ?? undefined\n}\n\nexport function isResponsesApiWebSearchEnabled(): boolean {\n  const config = getConfig()\n  return config.useResponsesApiWebSearch ?? true\n}\n\nexport function getClaudeTokenMultiplier(): number {\n  const config = getConfig()\n  return config.claudeTokenMultiplier ?? 1.15\n}\n","import consola from \"consola\"\nimport { getProxyForUrl } from \"proxy-from-env\"\nimport { Agent, ProxyAgent, setGlobalDispatcher, type Dispatcher } from \"undici\"\n\nlet proxyEnvDispatcher: Dispatcher | undefined\n\nexport function getProxyEnvDispatcher(): Dispatcher | undefined {\n  return proxyEnvDispatcher\n}\n\nexport function initProxyFromEnv(): void {\n  try {\n    const direct = new Agent()\n    const proxies = new Map<string, ProxyAgent>()\n\n    // We only need a minimal dispatcher that implements `dispatch` at runtime.\n    // Typing the object as `Dispatcher` forces TypeScript to require many\n    // additional methods. Instead, keep a plain object and cast when passing\n    // to `setGlobalDispatcher`.\n    const dispatcher = {\n      dispatch(\n        options: Dispatcher.DispatchOptions,\n        handler: Dispatcher.DispatchHandler,\n      ) {\n        try {\n          const origin =\n            typeof options.origin === \"string\" ?\n              new URL(options.origin)\n            : (options.origin as URL)\n          const get = getProxyForUrl as unknown as (\n            u: string,\n          ) => string | undefined\n          const raw = get(origin.toString())\n          const proxyUrl = raw && raw.length > 0 ? raw : undefined\n          if (!proxyUrl) {\n            consola.debug(`HTTP proxy bypass: ${origin.hostname}`)\n            return (direct as unknown as Dispatcher).dispatch(options, handler)\n          }\n          let agent = proxies.get(proxyUrl)\n          if (!agent) {\n            agent = new ProxyAgent(proxyUrl)\n            proxies.set(proxyUrl, agent)\n          }\n          let label = proxyUrl\n          try {\n            const u = new URL(proxyUrl)\n            label = `${u.protocol}//${u.host}`\n          } catch {\n            /* noop */\n          }\n          consola.debug(`HTTP proxy route: ${origin.hostname} via ${label}`)\n          return (agent as unknown as Dispatcher).dispatch(options, handler)\n        } catch {\n          return (direct as unknown as Dispatcher).dispatch(options, handler)\n        }\n      },\n      close() {\n        return direct.close()\n      },\n      destroy() {\n        return direct.destroy()\n      },\n    }\n\n    proxyEnvDispatcher = dispatcher as unknown as Dispatcher\n\n    if (typeof Bun !== \"undefined\") {\n      consola.debug(\"WebSocket proxy configured from environment (per-URL)\")\n      return\n    }\n\n    setGlobalDispatcher(proxyEnvDispatcher)\n    consola.debug(\"HTTP proxy configured from environment (per-URL)\")\n  } catch (err) {\n    consola.debug(\"Proxy setup skipped:\", err)\n  }\n}\n"],"mappings":";;;;;;;AA6DA,MAAM,wBAAwB;;;;;;AAO9B,MAAM,uBAAuB;;;;;;;;;;;;;;;;;;;AAoB7B,MAAM,gBAA2B;CAC/B,MAAM,EACJ,SAAS,EAAE,EACZ;CACD,WAAW,EAAE;CACb,eAAe,EAAE;CACjB,cAAc;EACZ,cAAc;EACd,iBAAiB;EACjB,gBAAgB;EAChB,WAAW;EACX,WAAW;EACZ;CACD,YAAY;CACZ,qCAAqC,EAAE;CACvC,uBAAuB;EACrB,cAAc;EACd,iBAAiB;EACjB,gBAAgB;EAChB,WAAW;EACX,WAAW;EACZ;CACD,gBAAgB;CAChB,0BAA0B;CAC1B,0BAA0B;CAC3B;AAED,IAAI,eAAiC;AAErC,SAAS,qBAAqB,aAAqC;CACjE,IAAI,OAAO,gBAAgB,UAAU;EACnC,IAAI,gBAAgB,KAAA,GAClB,QAAQ,KACN,gEACD;EAEH,OAAO;;CAGT,MAAM,wBAAwB,YAAY,MAAM;CAChD,IAAI,CAAC,uBAAuB;EAC1B,QAAQ,KACN,gEACD;EACD,OAAO;;CAGT,OAAO;;AAGT,SAAS,sBAA8B;CACrC,OAAO,YAAY,GAAG,CAAC,SAAS,MAAM;;AAGxC,SAAS,YAAY,OAAgD;CACnE,OAAO,iBAAiB,SAAS,UAAU;;AAG7C,SAAS,mBAAyB;CAChC,IAAI;EACF,GAAG,WAAW,MAAM,aAAa,GAAG,UAAU,OAAO,GAAG,UAAU,KAAK;SACjE;EACN,GAAG,UAAU,MAAM,SAAS,EAAE,WAAW,MAAM,CAAC;EAChD,GAAG,cACD,MAAM,aACN,GAAG,KAAK,UAAU,eAAe,MAAM,EAAE,CAAC,KAC1C,OACD;EACD,IAAI;GACF,GAAG,UAAU,MAAM,aAAa,IAAM;UAChC;GACN;;;;AAKN,SAAS,qBAAgC;CACvC,kBAAkB;CAClB,IAAI;EACF,MAAM,MAAM,GAAG,aAAa,MAAM,aAAa,OAAO;EACtD,IAAI,CAAC,IAAI,MAAM,EAAE;GACf,GAAG,cACD,MAAM,aACN,GAAG,KAAK,UAAU,eAAe,MAAM,EAAE,CAAC,KAC1C,OACD;GACD,OAAO;;EAET,OAAO,KAAK,MAAM,IAAI;UACf,OAAO;EACd,QAAQ,MAAM,oDAAoD,MAAM;EACxE,OAAO;;;AAIX,SAAS,6BAAwC;CAC/C,IAAI;EACF,MAAM,MAAM,GAAG,aAAa,MAAM,aAAa,OAAO;EACtD,IAAI,CAAC,IAAI,MAAM,EACb,OAAO,EAAE;EAEX,OAAO,KAAK,MAAM,IAAI;UACf,OAAO;EACd,IAAI,YAAY,MAAM,IAAI,MAAM,SAAS,UACvC,OAAO,EAAE;EAEX,IAAI,iBAAiB,aACnB,MAAM,IAAI,MAAM,kCAAkC,MAAM,cAAc;EAExE,MAAM;;;AAIV,SAAS,kBAAkB,QAAyB;CAClD,GAAG,UAAU,MAAM,SAAS,EAAE,WAAW,MAAM,CAAC;CAChD,GAAG,cACD,MAAM,aACN,GAAG,KAAK,UAAU,QAAQ,MAAM,EAAE,CAAC,KACnC,OACD;;AAGH,SAAS,mBAAmB,QAG1B;CACA,MAAM,eAAe,OAAO,gBAAgB,EAAE;CAC9C,MAAM,sBAAsB,cAAc,gBAAgB,EAAE;CAC5D,MAAM,wBAAwB,OAAO,yBAAyB,EAAE;CAChE,MAAM,+BAA+B,cAAc,yBAAyB,EAAE;CAE9E,MAAM,2BAA2B,OAAO,KAAK,oBAAoB,CAAC,QAC/D,UAAU,CAAC,OAAO,OAAO,cAAc,MAAM,CAC/C;CAED,MAAM,+BAA+B,OAAO,KAC1C,6BACD,CAAC,QAAQ,UAAU,CAAC,OAAO,OAAO,uBAAuB,MAAM,CAAC;CAEjE,MAAM,wBAAwB,yBAAyB,SAAS;CAChE,MAAM,4BAA4B,6BAA6B,SAAS;CAExE,IAAI,CAAC,yBAAyB,CAAC,2BAC7B,OAAO;EAAE,cAAc;EAAQ,SAAS;EAAO;CAGjD,OAAO;EACL,cAAc;GACZ,GAAG;GACH,cAAc;IACZ,GAAG;IACH,GAAG;IACJ;GACD,uBAAuB;IACrB,GAAG;IACH,GAAG;IACJ;GACF;EACD,SAAS;EACV;;AAGH,SAAS,kBAAkB,QAGzB;CACA,MAAM,wBAAwB,qBAAqB,OAAO,MAAM,YAAY;CAC5E,IAAI,uBAAuB;EACzB,IAAI,OAAO,MAAM,gBAAgB,uBAC/B,OAAO;GAAE,cAAc;GAAQ,SAAS;GAAO;EAGjD,OAAO;GACL,cAAc;IACZ,GAAG;IACH,MAAM;KACJ,GAAI,OAAO,QAAQ,EAAE;KACrB,aAAa;KACd;IACF;GACD,SAAS;GACV;;CAGH,MAAM,iBAAiB,4BAA4B;CACnD,MAAM,EAAE,iBAAiB,mBAAmB;EAC1C,GAAG;EACH,MAAM;GACJ,GAAI,eAAe,QAAQ,EAAE;GAC7B,aAAa,qBAAqB;GACnC;EACF,CAAC;CAEF,OAAO;EAAE;EAAc,SAAS;EAAM;;AAGxC,SAAgB,0BAAqC;CAEnD,MAAM,EAAE,cAAc,YAAY,mBADnB,oBAC4C,CAAC;CAC5D,MAAM,EACJ,cAAc,6BACd,SAAS,uBACP,kBAAkB,aAAa;CAGnC,IAF4B,WAAW,oBAGrC,IAAI;EACF,kBAAkB,4BAA4B;UACvC,YAAY;EACnB,IAAI,oBACF,MAAM;EAGR,QAAQ,KACN,sDACA,WACD;;CAIL,eAAe;CACf,OAAO;;AAGT,SAAgB,YAAuB;CACrC,iBAAiB,mBAAmB,oBAAoB,CAAC,CAAC;CAC1D,OAAO;;AAGT,SAAgB,eAA0B;CACxC,OAAO,yBAAyB;;AAGlC,SAAgB,uBAAuB,OAAuB;CAE5D,OADe,WACF,CAAC,eAAe,UAAU;;AAGzC,SAAgB,mBAA2C;CAEzD,MAAM,gBADS,WACa,CAAC;CAC7B,IAAI,CAAC,eACH,OAAO,EAAE,GAAI,cAAc,iBAAiB,EAAE,EAAG;CAGnD,MAAM,gBAAwC,EAAE;CAChD,KAAK,MAAM,CAAC,aAAa,gBAAgB,OAAO,QAAQ,cAAc,EAAE;EACtE,IACE,CAAC,eACE,OAAO,gBAAgB,YACvB,YAAY,WAAW,GAE1B;EAEF,cAAc,eAAe;;CAG/B,OAAO;;AAGT,SAAS,sBACP,eACwB;CACxB,MAAM,oBAA4C,EAAE;CACpD,KAAK,MAAM,CAAC,aAAa,gBAAgB,OAAO,QAAQ,cAAc,EAAE;EACtE,IAAI,CAAC,eAAe,CAAC,aACnB,MAAM,IAAI,MACR,kEACD;EAEH,kBAAkB,eAAe;;CAGnC,OAAO;;AAGT,SAAgB,iBACd,eACwB;CAMxB,kBAAkB;EAJhB,GAAG,4BAA4B;EAC/B,eAAe,sBAAsB,cAAc;EAGzB,CAAC;CAC7B,eAAe,cAAc;CAC7B,OAAO,kBAAkB;;AAG3B,SAAgB,mBAAmB,OAAuB;CACxD,OAAO,kBAAkB,CAAC,UAAU;;AAGtC,SAAgB,gBAAwB;CAEtC,OADe,WACF,CAAC,cAAc;;AAG9B,SAAgB,yCAAwD;CAEtE,OADe,WAEP,CAAC,uCACJ,cAAc,uCACd,EAAE;;AAIT,SAAgB,qCAAqC,OAAwB;CAC3E,OAAO,wCAAwC,CAAC,SAAS,MAAM;;AAGjE,SAAgB,2BACd,OAC0D;CAE1D,OADe,WACF,CAAC,wBAAwB,UAAU;;AAGlD,SAAgB,yBAAyB,KAAqB;CAC5D,OAAO,IAAI,MAAM,CAAC,QAAQ,SAAS,GAAG;;AAGxC,SAAS,2BACP,cACkB;CAClB,OAAO,iBAAiB,sBAAsB,kBAAkB;;AAGlE,SAAgB,wBACd,cACA,UACA,cACkB;CAClB,IAAI,aAAa,KAAA,GACf,OAAO,2BAA2B,aAAa;CAGjD,IAAI,aAAa,aACf,OAAO;CAGT,IAAI,aAAa,iBACf,OAAO;CAGT,QAAQ,KACN,YAAY,aAAa,yBAAyB,SAAS,qBAAqB,2BAA2B,aAAa,GACzH;CACD,OAAO,2BAA2B,aAAa;;AAGjD,SAAgB,kBAAkB,MAA6C;CAC7E,MAAM,eAAe,KAAK,MAAM;CAChC,IAAI,CAAC,cACH,OAAO;CAIT,MAAM,WADS,WACQ,CAAC,YAAY;CACpC,IAAI,CAAC,UACH,OAAO;CAGT,IAAI,SAAS,YAAY,OACvB,OAAO;CAGT,MAAM,OAAO,SAAS,QAAQ;CAC9B,IAAI,SAAS,eAAe,SAAS,qBAAqB;EACxD,QAAQ,KACN,YAAY,aAAa,4BAA4B,KAAK,oBAC3D;EACD,OAAO;;CAGT,MAAM,UAAU,yBAAyB,SAAS,WAAW,GAAG;CAChE,MAAM,UAAU,SAAS,UAAU,IAAI,MAAM;CAC7C,MAAM,WAAW,wBACf,cACA,SAAS,UACT,KACD;CACD,IAAI,CAAC,WAAW,CAAC,QAAQ;EACvB,QAAQ,KACN,YAAY,aAAa,2CAC1B;EACD,OAAO;;CAGT,OAAO;EACL,MAAM;EACN;EACA;EACA;EACA;EACA,QAAQ,SAAS;EACjB,mBAAmB,SAAS;EAC7B;;AASH,SAAgB,uBAAgC;CAE9C,OADe,WACF,CAAC,kBAAkB;;AAGlC,SAAgB,iCAA0C;CAExD,OADe,WACF,CAAC,4BAA4B;;AAG5C,SAAgB,qBAAyC;CAEvD,OADe,WACF,CAAC,mBAAmB,QAAQ,IAAI,qBAAqB,KAAA;;AAGpE,SAAgB,iCAA0C;CAExD,OADe,WACF,CAAC,4BAA4B;;AAG5C,SAAgB,2BAAmC;CAEjD,OADe,WACF,CAAC,yBAAyB;;;;AChgBzC,IAAI;AAEJ,SAAgB,wBAAgD;CAC9D,OAAO;;AAGT,SAAgB,mBAAyB;CACvC,IAAI;EACF,MAAM,SAAS,IAAI,OAAO;EAC1B,MAAM,0BAAU,IAAI,KAAyB;EAmD7C,qBAAqB;GA5CnB,SACE,SACA,SACA;IACA,IAAI;KACF,MAAM,SACJ,OAAO,QAAQ,WAAW,WACxB,IAAI,IAAI,QAAQ,OAAO,GACtB,QAAQ;KAIb,MAAM,MAAMA,eAAI,OAAO,UAAU,CAAC;KAClC,MAAM,WAAW,OAAO,IAAI,SAAS,IAAI,MAAM,KAAA;KAC/C,IAAI,CAAC,UAAU;MACb,QAAQ,MAAM,sBAAsB,OAAO,WAAW;MACtD,OAAQ,OAAiC,SAAS,SAAS,QAAQ;;KAErE,IAAI,QAAQ,QAAQ,IAAI,SAAS;KACjC,IAAI,CAAC,OAAO;MACV,QAAQ,IAAI,WAAW,SAAS;MAChC,QAAQ,IAAI,UAAU,MAAM;;KAE9B,IAAI,QAAQ;KACZ,IAAI;MACF,MAAM,IAAI,IAAI,IAAI,SAAS;MAC3B,QAAQ,GAAG,EAAE,SAAS,IAAI,EAAE;aACtB;KAGR,QAAQ,MAAM,qBAAqB,OAAO,SAAS,OAAO,QAAQ;KAClE,OAAQ,MAAgC,SAAS,SAAS,QAAQ;YAC5D;KACN,OAAQ,OAAiC,SAAS,SAAS,QAAQ;;;GAGvE,QAAQ;IACN,OAAO,OAAO,OAAO;;GAEvB,UAAU;IACR,OAAO,OAAO,SAAS;;GAII;EAE/B,IAAI,OAAO,QAAQ,aAAa;GAC9B,QAAQ,MAAM,wDAAwD;GACtE;;EAGF,oBAAoB,mBAAmB;EACvC,QAAQ,MAAM,mDAAmD;UAC1D,KAAK;EACZ,QAAQ,MAAM,wBAAwB,IAAI"}

package/dist/{server-S8A-8fWL.js → server-FPXzFkg9.js}

@@ -1,7 +1,7 @@
 import { t as PATHS } from "./paths-DC-mqCY3.js";
 import { D as generateTraceId, E as prepareMessageProxyHeaders, I as state, M as compactMessageSections, O as requestContext, P as compactSystemPromptStarts, T as prepareInteractionHeaders, _ as copilotWebSocketHeaders, c as getUUID, d as sleep, f as getCopilotUsage, g as copilotHeaders, h as copilotBaseUrl, j as compactAutoContinuePromptStarts, k as resolveTraceId$1, l as isNullish, m as forwardError, n as cacheModels, o as generateRequestIdFromPayload, p as HTTPError, s as getRootSessionId, u as parseUserIdMetadata, w as prepareForCompact } from "./utils-jHLgqAq2.js";
 import { a as isDeferredToolName, c as parseMcpToolSearchSentinel, d as shouldEnableResponsesToolSearch, i as isBridgeToolSearchName, l as resolveBridgeToolSearchName, o as listDeferredToolNames, r as formatToolSearchBridgeArguments, s as normalizeToolSearchBridgeArguments, t as BRIDGE_TOOL_SEARCH_NAME, u as selectDeferredToolsByNames } from "./tool-search-D3SN0jX-.js";
-import { _ as setModelMappings, a as getConfig, c as getProviderConfig, d as isMessagesApiEnabled, f as isResponsesApiContextManagementModel, g as resolveMappedModel, i as getClaudeTokenMultiplier, l as getReasoningEffortForModel, m as isResponsesApiWebSocketEnabled, o as getExtraPromptForModel, p as isResponsesApiWebSearchEnabled, r as getAnthropicApiKey, s as getModelMappings, t as getProxyEnvDispatcher, u as getSmallModel } from "./proxy-CSO5SLia.js";
+import { _ as setModelMappings, a as getConfig, c as getProviderConfig, d as isMessagesApiEnabled, f as isResponsesApiContextManagementModel, g as resolveMappedModel, i as getClaudeTokenMultiplier, l as getReasoningEffortForModel, m as isResponsesApiWebSocketEnabled, o as getExtraPromptForModel, p as isResponsesApiWebSearchEnabled, r as getAnthropicApiKey, s as getModelMappings, t as getProxyEnvDispatcher, u as getSmallModel } from "./proxy-DvlF9a-7.js";
 import consola from "consola";
 import fs from "node:fs/promises";
 import path from "node:path";
@@ -2343,6 +2343,25 @@ const applyModelDefaults = (payload, modelConfig) => {
 const applyMissingExtraBody = (payload, options) => {
 	for (const [key, value] of Object.entries(options.extraBody ?? {})) if (!Object.hasOwn(payload, key)) payload[key] = value;
 };
+const getRequestThinkingBudget = (payload) => {
+	const budget = payload.thinking?.budget_tokens;
+	if (typeof budget !== "number" || !Number.isFinite(budget)) return;
+	return budget;
+};
+const applyOpenAICompatibleThinkingBudget = (payload, source) => {
+	const thinkingBudget = getRequestThinkingBudget(source);
+	if (thinkingBudget !== void 0) {
+		payload.thinking_budget = thinkingBudget;
+		return;
+	}
+	if (payload.thinking_budget === void 0) delete payload.thinking_budget;
+};
+const applyOpenAICompatibleExtraBodyThinkingBudget = (payload, options) => {
+	const { extraBody } = options;
+	if (!extraBody || !Object.hasOwn(extraBody, "thinking_budget")) return;
+	const rawPayload = payload;
+	rawPayload.thinking_budget = extraBody.thinking_budget;
+};
 const handleOpenAICompatibleProviderMessages = async (c, options) => {
 	const { modelConfig, payload, provider, providerConfig } = options;
 	const openAIPayload = createOpenAICompatiblePayload(payload, modelConfig);
@@ -2373,6 +2392,7 @@ const createOpenAICompatiblePayload = (payload, modelConfig) => {
 		supportPdf: modelConfig?.supportPdf,
 		toolContentSupportType: modelConfig?.toolContentSupportType ?? []
 	});
+	applyOpenAICompatibleThinkingBudget(openAIPayload, payload);
 	if (payload.top_k !== void 0) openAIPayload.top_k = payload.top_k;
 	if (openAIPayload.stream) openAIPayload.stream_options = { include_usage: true };
 	normalizeOpenAICompatibleReasoningContent(openAIPayload);
@@ -2381,6 +2401,7 @@ const createOpenAICompatiblePayload = (payload, modelConfig) => {
 		source: payload
 	});
 	applyMissingExtraBody(openAIPayload, { extraBody: modelConfig?.extraBody });
+	applyOpenAICompatibleExtraBodyThinkingBudget(openAIPayload, { extraBody: modelConfig?.extraBody });
 	if (!Object.hasOwn(openAIPayload, "parallel_tool_calls")) openAIPayload.parallel_tool_calls = true;
 	if (modelConfig?.contextCache !== false) applyOpenAICompatibleContextCache(openAIPayload);
 	return openAIPayload;
@@ -4131,7 +4152,7 @@ const mergeContentWithText = (tr, textBlock) => {
 	if (hasToolRef(tr)) return tr;
 	return {
 		...tr,
-		content: [...tr.content, textBlock]
+		content: [...tr.content, stripContentBlockCacheControl(textBlock)]
 	};
 };
 const mergeContentWithTexts = (tr, textBlocks) => {
@@ -4145,22 +4166,29 @@ const mergeContentWithTexts = (tr, textBlocks) => {
 	if (hasToolRef(tr)) return tr;
 	return {
 		...tr,
-		content: [...tr.content, ...textBlocks]
+		content: [...tr.content, ...textBlocks.map(stripContentBlockCacheControl)]
 	};
 };
 const mergeContentWithAttachments = (tr, attachments) => {
+	const cleanAttachments = attachments.map(stripContentBlockCacheControl);
 	if (typeof tr.content === "string") return {
 		...tr,
 		content: [{
 			type: "text",
 			text: tr.content
-		}, ...attachments]
+		}, ...cleanAttachments]
 	};
 	return {
 		...tr,
-		content: [...tr.content, ...attachments]
+		content: [...tr.content, ...cleanAttachments]
 	};
 };
+const stripContentBlockCacheControl = (block) => {
+	if (!Object.hasOwn(block, "cache_control")) return block;
+	const copy = { ...block };
+	delete copy.cache_control;
+	return copy;
+};
 const isAttachmentBlock = (block) => {
 	return block.type === "image" || block.type === "document";
 };
@@ -4808,7 +4836,6 @@ const logger$1 = createHandlerLogger("responses-handler");
 const responsesHandlerDependencies = {
 	checkRateLimit,
 	createResponses,
-	getConfig,
 	isResponsesApiWebSearchEnabled
 };
 const handleResponses = async (c) => {
@@ -4824,7 +4851,6 @@ const handleResponses = async (c) => {
 		fallbackSessionId: sessionId,
 		model: payload.model
 	});
-	useFunctionApplyPatch(payload);
 	removeUnsupportedTools(payload);
 	if (!responsesHandlerDependencies.isResponsesApiWebSearchEnabled()) removeWebSearchTool(payload);
 	compactInputByLatestCompaction(payload);
@@ -4882,31 +4908,6 @@ const parseResponsesStreamEvent = (chunk) => {
 		return null;
 	}
 };
-const useFunctionApplyPatch = (payload) => {
-	if (responsesHandlerDependencies.getConfig().useFunctionApplyPatch ?? true) {
-		logger$1.debug("Using function tool apply_patch for responses");
-		if (Array.isArray(payload.tools)) {
-			const toolsArr = payload.tools;
-			for (let i = 0; i < toolsArr.length; i++) {
-				const t = toolsArr[i];
-				if (t.type === "custom" && t.name === "apply_patch") toolsArr[i] = {
-					type: "function",
-					name: t.name,
-					description: "Use the `apply_patch` tool to edit files",
-					parameters: {
-						type: "object",
-						properties: { input: {
-							type: "string",
-							description: "The entire contents of the apply_patch command"
-						} },
-						required: ["input"]
-					},
-					strict: false
-				};
-			}
-		}
-	}
-};
 const removeWebSearchTool = (payload) => {
 	if (!Array.isArray(payload.tools) || payload.tools.length === 0) return;
 	payload.tools = payload.tools.filter((t) => {
@@ -5039,4 +5040,4 @@ server.route("/:provider/v1/models", providerModelRoutes);
 //#endregion
 export { server };
 
-//# sourceMappingURL=server-S8A-8fWL.js.map
+//# sourceMappingURL=server-FPXzFkg9.js.map