@cortexkit/pi-anthropic-auth 2.4.5 → 2.4.6

package/dist/stream.js

@@ -1,4 +1,4 @@
-import { applyClaudeCodeHeaders, applyCustomHeaders, CACHE_KEEP_EXTENDED_TTL_BETA, CacheKeepManager, CacheKeepSessionRegistry, createProviderErrorResponseScope, createSseFramer, createStickyNoRouteResponse, decideStickyQuotaFailure, decodeSseFrame, dumpDirectRequest, FAST_MODE_BETA, getCache1hPersistentMode, getClaudeCodeIdentityForVerifiedAccount, getDefaultCacheKeepRegistryDirectory, getFallbackReauthLabels, getQuotaCheckIntervalMs, getRequestGzipLatch, getRequestGzipLatchPath, getRoutingMode, getScopedQuotaWindowForModel, getThinkingPrefixMismatchBehavior, isCache1hPersistentlyEnabled, isCacheKeepHybridActive, isDumpPersistentlyEnabled, isFastModePersistentlyEnabled, isKillswitchEnabled, isNativeTransientAccountError, isOAuthAccount, isPermanentRefreshError, isValidApiBaseURL, killswitchPassesPolicy, logger, mergeAnthropicBetas, NativeAccountRefusal, NativeVaultNoAccountError, nativeAccountCanServe, parseSseFrame, QuotaManager, quotaBackoffActive, quotaSnapshotHasStandardWindows, quotaSnapshotModelScopeIsExhausted, quotaSnapshotPassesModelScope, quotaSnapshotPassesPolicy, STICKY_ROUTING_MAIN_ACCOUNT_ID, StickySessionRouter, sendViaRelay, sendWithRequestGzip, setDumpEnabled, shouldFallbackStatus, stickyQuotaSnapshotIsFresh, stickyRetryAfterWithJitter, stickyRouteFamilyForModel, } from '@cortexkit/anthropic-auth-core';
+import { applyClaudeCodeHeaders, applyCustomHeaders, CACHE_KEEP_EXTENDED_TTL_BETA, CacheKeepManager, CacheKeepSessionRegistry, createProviderErrorResponseScope, createSseFramer, createStickyNoRouteResponse, decideStickyQuotaFailure, decodeSseFrame, dumpDirectRequest, FAST_MODE_BETA, getCache1hPersistentMode, getClaudeCodeIdentityForVerifiedAccount, getDefaultCacheKeepRegistryDirectory, getFallbackReauthLabels, getQuotaCheckIntervalMs, getRequestGzipLatch, getRequestGzipLatchPath, getRoutingMode, getScopedQuotaWindowForModel, getThinkingPrefixMismatchBehavior, isCache1hPersistentlyEnabled, isCacheKeepHybridActive, isDumpPersistentlyEnabled, isFastModePersistentlyEnabled, isKillswitchEnabled, isNativeTransientAccountError, isOAuthAccount, isPermanentRefreshError, isValidApiBaseURL, killswitchPassesPolicy, logger, mergeAnthropicBetas, NativeAccountRefusal, NativeVaultNoAccountError, nativeAccountCanServe, parseSseFrame, QuotaManager, quotaBackoffActive, quotaExhaustionFromHeaders, quotaSnapshotHasStandardWindows, quotaSnapshotModelScopeIsExhausted, quotaSnapshotPassesModelScope, quotaSnapshotPassesPolicy, STICKY_ROUTING_MAIN_ACCOUNT_ID, StickySessionRouter, sendViaRelay, sendWithRequestGzip, setDumpEnabled, shouldFallbackStatus, stickyQuotaSnapshotIsFresh, stickyRetryAfterWithJitter, stickyRouteFamilyForModel, } from '@cortexkit/anthropic-auth-core';
 import { calculateCost, createAssistantMessageEventStream, } from '@earendil-works/pi-ai';
 import { buildAnthropicRequest, fromClaudeCodeToolName } from './convert.js';
 import { getPiNativeRuntime, piAttemptAccessToken, } from './native.js';
@@ -1051,23 +1051,26 @@ async function executeWithFallback(options) {
             let migrate = (preflight instanceof Response && preflight.status === 403) ||
                 permanentAuthFailure;
             if (primaryResponseAllowsApiFallback(preflight)) {
-                let quota;
-                try {
-                    quota =
-                        route.id === STICKY_ROUTING_MAIN_ACCOUNT_ID
-                            ? await quotaManager.refreshMain(mainAccountId, '')
-                            : await quotaManager.refreshFallback(route.id, '', route.account);
-                }
-                catch {
-                    // Retain affinity when the quota probe itself is unavailable.
-                    quota = undefined;
+                const responseHeaders = preflight instanceof Response ? preflight.headers : response.headers;
+                let quota = quotaExhaustionFromHeaders(responseHeaders, options.model.id);
+                // Do not extend the usage endpoint's own rate limit when the refusal
+                // already confirms exhaustion.
+                if (Object.keys(quota).length === 0) {
+                    try {
+                        quota =
+                            route.id === STICKY_ROUTING_MAIN_ACCOUNT_ID
+                                ? await quotaManager.refreshMain(mainAccountId, '')
+                                : await quotaManager.refreshFallback(route.id, '', route.account);
+                    }
+                    catch {
+                        // Retain affinity when the quota probe itself is unavailable.
+                        quota = undefined;
+                    }
                 }
                 const decision = decideStickyQuotaFailure({
                     quota,
                     modelId: options.model.id,
-                    responseHeaders: preflight instanceof Response
-                        ? preflight.headers
-                        : response.headers,
+                    responseHeaders,
                 });
                 if (decision.action === 'hold') {
                     const headers = new Headers(preflight instanceof Response

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@cortexkit/pi-anthropic-auth",
-  "version": "2.4.5",
+  "version": "2.4.6",
   "type": "module",
   "repository": {
     "type": "git",
@@ -36,7 +36,7 @@
     "prepublishOnly": "bun run build"
   },
   "dependencies": {
-    "@cortexkit/anthropic-auth-core": "2.4.5"
+    "@cortexkit/anthropic-auth-core": "2.4.6"
   },
   "peerDependencies": {
     "@earendil-works/pi-ai": ">=0.86.1 <2",