feat: adaptive reasoning detection for Codex Chat providers

Auto-detect each Chat-routed Codex provider's reasoning interface from
its name, base URL, and model, then inject the matching thinking
parameter without manual configuration:

- Platform-first inference (OpenRouter, SiliconFlow) overrides model
  rules, since the same model exposes different reasoning controls
  depending on the hosting platform.
- Effort tiers are forwarded only to providers that support them
  (DeepSeek, OpenRouter, and StepFun's step-3.5-flash-2603); on/off-only
  providers (Kimi, GLM, Qwen, MiniMax, MiMo, SiliconFlow) drop the level
  instead of sending a field the upstream rejects.
- OpenRouter uses the native reasoning:{effort} object, clamps max to
  xhigh (its enum has no max), and forwards an explicit effort:"none" so
  reasoning can be turned off.
- StepFun falls back to inference so per-model effort support is honored
  (the static preset would have forced effort on step-3.5-flash too).

Includes the Codex provider-form reasoning controls, i18n strings
(zh/en/ja), and response-side reasoning extraction.

src/components/providers/forms/ProviderForm.tsx

@@ -16,6 +16,7 @@ import type {
   ClaudeApiFormat,
   CodexApiFormat,
   CodexCatalogModel,
+  CodexChatReasoning,
   ClaudeApiKeyField,
 } from "@/types";
 import {
@@ -174,6 +175,41 @@ export const normalizeCodexCatalogModelsForSave = (
   return normalized;
 };
 
+const normalizeCodexChatReasoningForSave = (
+  value?: CodexChatReasoning,
+): CodexChatReasoning | undefined => {
+  const supportsEffort = value?.supportsEffort === true;
+  const supportsThinking = value?.supportsThinking === true || supportsEffort;
+  const hasExplicitConfig = value && Object.keys(value).length > 0;
+
+  if (!supportsThinking && !supportsEffort) {
+    return hasExplicitConfig
+      ? {
+          supportsThinking: false,
+          supportsEffort: false,
+          thinkingParam: "none",
+          effortParam: "none",
+          outputFormat: value?.outputFormat ?? "auto",
+        }
+      : undefined;
+  }
+
+  return {
+    supportsThinking,
+    supportsEffort,
+    thinkingParam: supportsThinking
+      ? (value?.thinkingParam ?? "thinking")
+      : "none",
+    effortParam: supportsEffort
+      ? (value?.effortParam ?? "reasoning_effort")
+      : "none",
+    effortValueMode: supportsEffort
+      ? (value?.effortValueMode ?? "passthrough")
+      : undefined,
+    outputFormat: value?.outputFormat ?? "auto",
+  };
+};
+
 export interface ProviderFormProps {
   appId: AppId;
   providerId?: string;
@@ -316,6 +352,7 @@ function ProviderFormFull({
         initialData?.meta?.pricingModelSource,
       ),
     });
+    setCodexChatReasoning(initialData?.meta?.codexChatReasoning ?? {});
   }, [appId, initialData, supportsFullUrl]);
 
   const defaultValues: ProviderFormData = useMemo(
@@ -466,6 +503,10 @@ function ProviderFormFull({
   const [codexFastMode, setCodexFastMode] = useState<boolean>(
     () => initialData?.meta?.codexFastMode ?? false,
   );
+  const [codexChatReasoning, setCodexChatReasoning] =
+    useState<CodexChatReasoning>(
+      () => initialData?.meta?.codexChatReasoning ?? {},
+    );
 
   const {
     codexAuth,
@@ -530,6 +571,7 @@ function ProviderFormFull({
     if (appId === "codex" && !initialData && selectedPresetId === "custom") {
       const template = getCodexCustomTemplate();
       resetCodexConfig(template.auth, template.config);
+      setCodexChatReasoning({});
     }
   }, [appId, initialData, selectedPresetId, resetCodexConfig]);
 
@@ -1336,6 +1378,12 @@ function ProviderFormFull({
           ? selectedGitHubAccountId
           : undefined,
       codexFastMode: isCodexOauthProvider ? codexFastMode : undefined,
+      codexChatReasoning:
+        appId === "codex" &&
+        category !== "official" &&
+        localCodexApiFormat === "openai_chat"
+          ? normalizeCodexChatReasoningForSave(codexChatReasoning)
+          : undefined,
       testConfig: testConfig.enabled ? testConfig : undefined,
       costMultiplier: pricingConfig.enabled
         ? pricingConfig.costMultiplier
@@ -1473,6 +1521,7 @@ function ProviderFormFull({
       if (appId === "codex") {
         const template = getCodexCustomTemplate();
         resetCodexConfig(template.auth, template.config);
+        setCodexChatReasoning({});
         setLocalCodexApiFormat(
           codexApiFormatFromWireApi(extractCodexWireApi(template.config)) ??
             "openai_responses",
@@ -1513,6 +1562,7 @@ function ProviderFormFull({
       const config = preset.config ?? "";
 
       resetCodexConfig(auth, config, preset.modelCatalog ?? []);
+      setCodexChatReasoning(preset.codexChatReasoning ?? {});
       setLocalCodexApiFormat(
         preset.apiFormat ??
           codexApiFormatFromWireApi(extractCodexWireApi(config)) ??
@@ -1984,6 +2034,8 @@ function ProviderFormFull({
               onAutoSelectChange={setEndpointAutoSelect}
               apiFormat={localCodexApiFormat}
               onApiFormatChange={handleCodexApiFormatChange}
+              codexChatReasoning={codexChatReasoning}
+              onCodexChatReasoningChange={setCodexChatReasoning}
               catalogModels={codexCatalogModels}
               onCatalogModelsChange={setCodexCatalogModels}
               speedTestEndpoints={speedTestEndpoints}