chore(ai-gateway): remove Cloudflare analytics

apps/web/src/app/api/openrouter/[...path]/route.test.ts

@@ -20,7 +20,6 @@ import {
   getOpenRouterModelsFromDatabase,
   isValidOpenRouterModelId,
 } from '@/lib/ai-gateway/providers/gateway-models-cache';
-import { emitApiMetricsForResponse } from '@/lib/ai-gateway/o11y/api-metrics.server';
 import { accountForMicrodollarUsage, INVALID_TOKEN_CODE } from '@/lib/ai-gateway/llm-proxy-helpers';
 import { ReasoningDetailsTransform, type Provider } from '@/lib/ai-gateway/providers/types';
 import { fetchEfficientAutoDecision } from '@/lib/ai-gateway/auto-routing-decision';
@@ -82,11 +81,6 @@ jest.mock('@/lib/ai-gateway/providers/direct-byok', () => ({
 }));
 jest.mock('@/lib/ai-gateway/providers/upstream-request');
 jest.mock('@/lib/ai-gateway/providers/gateway-models-cache');
-jest.mock('@/lib/ai-gateway/o11y/api-metrics.server', () => ({
-  emitApiMetricsForResponse: jest.fn(),
-  getToolsAvailable: jest.fn(() => false),
-  getToolsUsed: jest.fn(() => false),
-}));
 jest.mock('@/lib/ai-gateway/rewriteModelResponse', () => {
   const actual = jest.requireActual('@/lib/ai-gateway/rewriteModelResponse');
   const { wrapInSafeNextResponse } = jest.requireActual('@/lib/ai-gateway/llm-proxy-helpers');
@@ -133,7 +127,6 @@ const mockedGetProvider = jest.mocked(getProvider);
 const mockedUpstreamRequest = jest.mocked(upstreamRequest);
 const mockedGetOpenRouterModels = jest.mocked(getOpenRouterModelsFromDatabase);
 const mockedIsValidOpenRouterModelId = jest.mocked(isValidOpenRouterModelId);
-const mockedEmitApiMetricsForResponse = jest.mocked(emitApiMetricsForResponse);
 const mockedAccountForMicrodollarUsage = jest.mocked(accountForMicrodollarUsage);
 const mockedFetchEfficientAutoDecision = jest.mocked(fetchEfficientAutoDecision);
 const mockedCollectDeniedAutoRoutingModelIds = jest.mocked(collectDeniedAutoRoutingModelIds);
@@ -267,7 +260,6 @@ describe('POST /api/openrouter/v1/chat/completions bearer audiences', () => {
       type: 'success',
       response: upstreamJsonResponse({ id: 'chatcmpl-1', model: 'openai/gpt-4o', choices: [] }),
     });
-    mockedEmitApiMetricsForResponse.mockReturnValue(undefined);
     mockedAccountForMicrodollarUsage.mockReturnValue(undefined);
   });
 
@@ -568,7 +560,6 @@ describe('POST /api/openrouter/v1/chat/completions request handling', () => {
       type: 'success',
       response: upstreamJsonResponse({ id: 'chatcmpl-1', model: 'openai/gpt-4o', choices: [] }),
     });
-    mockedEmitApiMetricsForResponse.mockReturnValue(undefined);
     mockedAccountForMicrodollarUsage.mockReturnValue(undefined);
   });
 
@@ -916,7 +907,6 @@ describe('kilo-auto/efficient classifier billing', () => {
         choices: [],
       }),
     });
-    mockedEmitApiMetricsForResponse.mockReturnValue(undefined);
     mockedAccountForMicrodollarUsage.mockReturnValue(undefined);
     mockedLogMicrodollarUsage.mockResolvedValue(null);
     mockedGetEffectiveModelDecision.mockResolvedValue({ allowed: true });
@@ -1385,7 +1375,6 @@ describe('auto-routing shadow classifier', () => {
       type: 'success',
       response: upstreamJsonResponse({ id: 'chatcmpl-1', model: 'openai/gpt-4o', choices: [] }),
     });
-    mockedEmitApiMetricsForResponse.mockReturnValue(undefined);
     mockedAccountForMicrodollarUsage.mockReturnValue(undefined);
     mockedApplyResolvedAutoModel.mockImplementation(async (opts, request) => {
       if (opts.efficientDecision) await opts.efficientDecision();

apps/web/src/app/api/openrouter/[...path]/route.ts

@@ -77,7 +77,6 @@ import {
   checkPromotionLimit,
 } from '@/lib/free-model-rate-limiter';
 import { PROMOTION_MAX_REQUESTS, PROMOTION_WINDOW_HOURS } from '@/lib/constants';
-import { emitApiMetricsForResponse } from '@/lib/ai-gateway/o11y/api-metrics.server';
 import {
   gatewayRateLimitKey,
   isGatewayAccountRateLimited,
@@ -932,7 +931,7 @@ async function openRouterPost(request: NextRequest): Promise<NextResponseType<un
   }
   if (attempt.type === 'error') return attempt.response;
 
-  const { response, toolsAvailable, toolsUsed, experimentPromptCapture } = attempt;
+  const { response, experimentPromptCapture } = attempt;
   if (experimentPromptCapture) usageContext.experimentPromptCapture = experimentPromptCapture;
   const finalUpstreamModel = requestBodyParsed.body.model ?? effectiveModelIdLowerCased;
   logExceptInTest(
@@ -945,25 +944,6 @@ async function openRouterPost(request: NextRequest): Promise<NextResponseType<un
   const ttfbMs = Math.max(0, Math.round(performance.now() - requestStartedAt));
   usageContext.ttfb_ms = ttfbMs;
 
-  emitApiMetricsForResponse(
-    {
-      kiloUserId: user.id,
-      organizationId,
-      isAnonymous: isAnonymousContext(user),
-      isStreaming: requestBodyParsed.body.stream === true,
-      userByok: !!effectiveProviderContext.userByok,
-      mode: modeHeader || undefined,
-      provider: effectiveProviderContext.provider.id,
-      requestedModel: requestedModelLowerCased,
-      resolvedModel: normalizeModelId(effectiveModelIdLowerCased),
-      toolsAvailable,
-      toolsUsed,
-      ttfbMs,
-      statusCode: response.status,
-    },
-    response.clone(),
-    requestStartedAt
-  );
   usageContext.status_code = response.status;
 
   // Handle OpenRouter 402 errors - don't pass them through to the client. We need to pay, not them.

apps/web/src/app/api/openrouter/audio/transcriptions/route.test.ts

@@ -4,7 +4,6 @@ import { getBalanceAndOrgSettings } from '@/lib/organizations/organization-usage
 import { isFreeModel } from '@/lib/ai-gateway/is-free-model';
 import { isAutoTopUpInFlight } from '@/lib/autoTopUpInFlight';
 import type { User } from '@kilocode/db/schema';
-import { emitApiMetricsForResponse } from '@/lib/ai-gateway/o11y/api-metrics.server';
 import type { OrganizationSettings } from '@/lib/organizations/organization-types';
 
 jest.mock('next/server', () => {
@@ -17,9 +16,6 @@ jest.mock('next/server', () => {
 jest.mock('@/lib/user/server');
 jest.mock('@/lib/organizations/organization-usage');
 jest.mock('@/lib/autoTopUpInFlight');
-jest.mock('@/lib/ai-gateway/o11y/api-metrics.server', () => ({
-  emitApiMetricsForResponse: jest.fn(),
-}));
 jest.mock('@/lib/ai-gateway/is-free-model', () => ({
   isFreeModel: jest.fn(),
 }));
@@ -35,7 +31,6 @@ const mockedGetUserFromAuth = jest.mocked(getUserFromAuth);
 const mockedGetBalanceAndOrgSettings = jest.mocked(getBalanceAndOrgSettings);
 const mockedIsFreeModel = jest.mocked(isFreeModel);
 const mockedIsAutoTopUpInFlight = jest.mocked(isAutoTopUpInFlight);
-const mockedEmitApiMetricsForResponse = jest.mocked(emitApiMetricsForResponse);
 const mockedFetch = jest.fn() as jest.MockedFunction<typeof globalThis.fetch>;
 const originalFetch = globalThis.fetch;
 
@@ -152,9 +147,6 @@ describe('POST /api/gateway/v1/audio/transcriptions', () => {
     expect(upstream.input_audio).toEqual({ data: 'UklGRiQA', format: 'wav' });
     expect(upstream.safety_identifier).toBeTruthy();
     expect(upstream.user).toBe(upstream.safety_identifier);
-    expect(mockedEmitApiMetricsForResponse.mock.calls[0]?.[0]).not.toMatchObject({
-      feature: 'vscode-extension',
-    });
   });
 
   it('forwards organization provider policy through the OpenRouter provider field', async () => {

apps/web/src/app/api/openrouter/audio/transcriptions/route.ts

@@ -26,8 +26,6 @@ import type { OpenRouterProviderConfig } from '@/lib/ai-gateway/providers/openro
 import { ProxyErrorType } from '@/lib/proxy-error-types';
 import { getBalanceAndOrgSettings } from '@/lib/organizations/organization-usage';
 import { isFreeModel } from '@/lib/ai-gateway/is-free-model';
-import { emitApiMetricsForResponse } from '@/lib/ai-gateway/o11y/api-metrics.server';
-import { normalizeModelId } from '@/lib/ai-gateway/model-utils';
 import {
   buildUpstreamBody,
   extractTranscriptionPromptInfo,
@@ -343,25 +341,6 @@ export async function POST(request: NextRequest): Promise<NextResponseType<unkno
   usageContext.ttfb_ms = ttfbMs;
   usageContext.status_code = response.status;
 
-  emitApiMetricsForResponse(
-    {
-      kiloUserId: user.id,
-      organizationId,
-      isAnonymous: false,
-      isStreaming: false,
-      userByok: !!userByok,
-      provider: provider.id,
-      requestedModel: requestedModelLowerCased,
-      resolvedModel: normalizeModelId(requestedModelLowerCased),
-      toolsAvailable: [],
-      toolsUsed: [],
-      ttfbMs,
-      statusCode: response.status,
-    },
-    response.clone(),
-    requestStartedAt
-  );
-
   if (response.status === 402 && !userByok) {
     await captureProxyError({
       user,

apps/web/src/app/api/openrouter/embeddings/route.ts

@@ -33,8 +33,6 @@ import {
   isAnonymousContext,
   type AnonymousUserContext,
 } from '@/lib/anonymous';
-import { emitApiMetricsForResponse } from '@/lib/ai-gateway/o11y/api-metrics.server';
-import { normalizeModelId } from '@/lib/ai-gateway/model-utils';
 import {
   buildUpstreamBody,
   type EmbeddingProxyRequest,
@@ -313,25 +311,6 @@ export async function POST(request: NextRequest): Promise<NextResponseType<unkno
   const ttfbMs = Math.max(0, Math.round(performance.now() - requestStartedAt));
   usageContext.ttfb_ms = ttfbMs;
 
-  emitApiMetricsForResponse(
-    {
-      kiloUserId: user.id,
-      organizationId,
-      isAnonymous: isAnonymousContext(user),
-      isStreaming: false,
-      userByok: !!userByok,
-      provider: provider.id,
-      requestedModel: requestedModelLowerCased,
-      resolvedModel: normalizeModelId(requestedModelLowerCased),
-      toolsAvailable: [],
-      toolsUsed: [],
-      ttfbMs,
-      statusCode: response.status,
-    },
-    response.clone(),
-    requestStartedAt
-  );
-
   usageContext.status_code = response.status;
 
   // Handle upstream 402 — don't pass through to client (same as chat completions)