refactor: simplify AI moderation pipeline to Primary AI only, fix stickerCache executeGet

- Remove Groq and NVIDIA fallback from text moderation pipeline
- Remove GROQ_API_KEY, GROQ_MODERATION_*, NVIDIA_NEMOTRON_* config vars
- Update source enum: remove 'groq' and 'nvidia' from cache/schema types
- Rewrite indonesianTextNormalizer to use Primary AI only (no fallback chain)
- Rewrite remote test to test Primary AI only
- Fix stickerCache executeGet missing empty params array

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

tests/moderation/indonesianTextNormalizer.remote.test.ts

@@ -2,23 +2,9 @@ import { beforeEach, describe, expect, it, vi } from "vitest";
 import { config } from "../../src/config";
 
 const mocks = vi.hoisted(() => ({
-  axiosPost: vi.fn(),
   openaiCreate: vi.fn(),
 }));
 
-vi.mock("axios", () => ({
-  default: {
-    post: mocks.axiosPost,
-    isAxiosError: (error: unknown) =>
-      Boolean(
-        error &&
-          typeof error === "object" &&
-          "isAxiosError" in error &&
-          (error as { isAxiosError?: unknown }).isAxiosError,
-      ),
-  },
-}));
-
 vi.mock("openai", () => ({
   default: class MockOpenAI {
     chat = {
@@ -29,20 +15,13 @@ vi.mock("openai", () => ({
   },
 }));
 
-describe("detectIndonesianBadwords remote fallback", () => {
+describe("detectIndonesianBadwords primary AI", () => {
   beforeEach(() => {
-    mocks.axiosPost.mockReset();
     mocks.openaiCreate.mockReset();
-    config.NVIDIA_NEMOTRON_API_KEY = "test-nemotron-key";
     config.AI_LLM_API_KEY = "test-primary-key";
   });
 
-  it("falls back to primary AI after Nemotron rate limits and caches the result", async () => {
-    mocks.axiosPost.mockRejectedValue({
-      isAxiosError: true,
-      response: { status: 429 },
-      message: "Too Many Requests",
-    });
+  it("calls primary AI and caches the result", async () => {
     mocks.openaiCreate.mockResolvedValue({
       choices: [
         {
@@ -62,7 +41,17 @@ describe("detectIndonesianBadwords remote fallback", () => {
 
     expect(first).toEqual(["harassment"]);
     expect(second).toEqual(["harassment"]);
-    expect(mocks.axiosPost).toHaveBeenCalledTimes(1);
     expect(mocks.openaiCreate).toHaveBeenCalledTimes(1);
   });
+
+  it("returns empty array when primary AI fails", async () => {
+    mocks.openaiCreate.mockRejectedValue(new Error("API down"));
+
+    const { detectIndonesianBadwords } = await import(
+      "../../src/moderation/indonesianTextNormalizer"
+    );
+
+    const result = await detectIndonesianBadwords("some text");
+    expect(result).toEqual([]);
+  });
 });

tests/moderation/indonesianTextNormalizer.test.ts

@@ -7,14 +7,10 @@ import {
   normalizeDiscordCustomEmoji,
 } from "../../src/moderation/indonesianTextNormalizer";
 
-const originalNemotronKey = config.NVIDIA_NEMOTRON_API_KEY;
 const originalPrimaryAiKey = config.AI_LLM_API_KEY;
-const originalGroqKey = config.GROQ_API_KEY;
 
 function disableRemoteModeration(): void {
-  config.NVIDIA_NEMOTRON_API_KEY = undefined;
   config.AI_LLM_API_KEY = undefined;
-  config.GROQ_API_KEY = undefined;
 }
 
 disableRemoteModeration();
@@ -24,9 +20,7 @@ afterEach(() => {
 });
 
 afterAll(() => {
-  config.NVIDIA_NEMOTRON_API_KEY = originalNemotronKey;
   config.AI_LLM_API_KEY = originalPrimaryAiKey;
-  config.GROQ_API_KEY = originalGroqKey;
 });
 
 describe("normalizeDiscordCustomEmoji", () => {