feat(llm): add GPT-5.5 codex fallback

Author: r1n7aro

src/__tests__/modelStoreEffort.test.ts

@@ -35,16 +35,20 @@ describe("useModelStore OpenAI effort mapping", () => {
     modelServiceMocks.saveLastEffort.mockResolvedValue(undefined);
   });
 
-  it("uses GPT-5.4 as the Codex fallback catalog", () => {
+  it("includes GPT-5.5 and GPT-5.4 in the Codex fallback catalog", () => {
     const authStore = useAuthStore();
     authStore.codexAuthenticated = true;
     const modelStore = useModelStore();
 
+    expect(modelStore.codexModels.map((model) => model.id)).toEqual([
+      "openai/gpt-5.5",
+      "openai/gpt-5.4",
+    ]);
+    expect(modelStore.availableModels.some((model) => model.id === "openai/gpt-5.5")).toBe(true);
     expect(modelStore.availableModels.some((model) => model.id === "openai/gpt-5.4")).toBe(true);
-    expect(modelStore.availableModels.some((model) => model.id === "openai/gpt-5.5")).toBe(false);
   });
 
-  it("exposes GPT-5.5 only after the remote Codex catalog returns it", async () => {
+  it("uses the remote Codex catalog when it is available", async () => {
     const authStore = useAuthStore();
     authStore.codexAuthenticated = true;
     modelServiceMocks.getCodexAvailableModels.mockResolvedValue([

src/stores/model.ts

@@ -25,14 +25,23 @@ const builtinModels: ModelOption[] = [
 ];
 
 const codexFallbackModels: ModelOption[] = [
+  {
+    id: "openai/gpt-5.5",
+    name: "GPT-5.5",
+    provider: "openai_codex",
+    defaultEffort: "medium",
+    supportedEfforts: ["low", "medium", "high", "xhigh"],
+    additionalSpeedTiers: ["fast"],
+    isDefault: true,
+  },
   {
     id: "openai/gpt-5.4",
     name: "GPT-5.4",
     provider: "openai_codex",
     defaultEffort: "medium",
     supportedEfforts: ["low", "medium", "high", "xhigh"],
     additionalSpeedTiers: ["fast"],
-    isDefault: true,
+    isDefault: false,
   },
 ];