fix: giảm spam 429 từ Claude OAuth usage endpoint

- claudeAutoPing: cache resetAt in-mem, bỏ qua poll usage cho tới gần reset
- ProviderLimits: throttle auto-refresh Claude 3 phút, nút bấm tay vẫn refresh ngay
- claude.js: 429 ở OAuth usage → cooldown 3 phút, fallback legacy (không ảnh hưởng chat)

Co-authored-by: Cursor <cursoragent@cursor.com>

open-sse/services/usage/claude.js

@@ -14,11 +14,19 @@ const CLAUDE_CONFIG = {
   apiVersion: ANTHROPIC_API_VERSION,
 };
 
-/**
- * Claude Usage - Primary: OAuth endpoint, Fallback: legacy settings/org endpoint
- */
+// OAuth usage endpoint rate-limits (429); cool down per-token to stop hammering it.
+// Only the quota endpoint is affected — chat with the same token still works.
+const OAUTH_429_COOLDOWN_MS = 180000;
+const oauthCooldown = new Map();
+
 export async function getClaudeUsage(accessToken, proxyOptions = null) {
   try {
+    // Skip OAuth usage call while this token is cooling down from a recent 429
+    const cooldownUntil = oauthCooldown.get(accessToken);
+    if (cooldownUntil && Date.now() < cooldownUntil) {
+      return await getClaudeUsageLegacy(accessToken, proxyOptions);
+    }
+
     // Primary: OAuth usage endpoint (Claude Code consumer OAuth tokens)
     const oauthResponse = await proxyAwareFetch(CLAUDE_CONFIG.oauthUsageUrl, {
       method: "GET",
@@ -73,6 +81,11 @@ export async function getClaudeUsage(accessToken, proxyOptions = null) {
       };
     }
 
+    // Cool down OAuth usage polling after a 429 (quota endpoint only)
+    if (oauthResponse.status === 429) {
+      oauthCooldown.set(accessToken, Date.now() + OAUTH_429_COOLDOWN_MS);
+    }
+
     // Fallback: legacy settings + org usage endpoint
     console.warn(`[Claude Usage] OAuth endpoint returned ${oauthResponse.status}, falling back to legacy`);
     return await getClaudeUsageLegacy(accessToken, proxyOptions);