feat: comprehensive rate limit handling overhaul (inspired by opencode-antigravity-auth)

This commit addresses "Max retries exceeded" errors during stress testing where
all accounts would become exhausted simultaneously due to short per-second rate
limits triggering cascading failures.

## Rate Limit Parser (`rate-limit-parser.js`)
- Remove 2s buffer enforcement that caused cascading failures when API returned
  short reset times (200-600ms). Now adds 200ms buffer for sub-500ms resets
- Add `parseRateLimitReason()` for smart backoff based on error type:
  QUOTA_EXHAUSTED, RATE_LIMIT_EXCEEDED, MODEL_CAPACITY_EXHAUSTED, SERVER_ERROR

## Message/Streaming Handlers
- Add per-account+model rate limit state tracking with exponential backoff
- For short rate limits (< 1 second), wait and retry on same account instead
  of switching - prevents thundering herd when all accounts hit per-second limits
- Add throttle wait support for fallback modes (emergency/lastResort)
- Add `calculateSmartBackoff()` with progressive tiers by error type

## HybridStrategy (`hybrid-strategy.js`)
- Refactor `#getCandidates()` to return 4 fallback levels:
  - `normal`: All filters pass (health, tokens, quota)
  - `quota`: Bypass critical quota check
  - `emergency`: Bypass health check when ALL accounts unhealthy
  - `lastResort`: Bypass BOTH health AND token bucket checks
- Add throttle wait times: 500ms for lastResort, 250ms for emergency
- Fix LRU calculation to use seconds (matches opencode-antigravity-auth)

## Health Tracker
- Increase `recoveryPerHour` from 2 to 10 for faster recovery (1 hour vs 5 hours)

## Account Manager
- Add consecutive failure tracking: `getConsecutiveFailures()`,
  `incrementConsecutiveFailures()`, `resetConsecutiveFailures()`
- Add cooldown mechanism separate from rate limits with `CooldownReason`
- Reset consecutive failures on successful request

## Base Strategy
- Add `isAccountCoolingDown()` check in `isAccountUsable()`

## Constants
- Replace fixed `CAPACITY_RETRY_DELAY_MS` with progressive `CAPACITY_BACKOFF_TIERS_MS`
- Add `BACKOFF_BY_ERROR_TYPE` for smart backoff
- Add `QUOTA_EXHAUSTED_BACKOFF_TIERS_MS` for progressive quota backoff
- Add `MIN_BACKOFF_MS` floor to prevent "Available in 0s" loops
- Increase `MAX_CAPACITY_RETRIES` from 3 to 5
- Reduce `RATE_LIMIT_DEDUP_WINDOW_MS` from 5s to 2s

## Frontend
- Remove `capacityRetryDelayMs` config (replaced by progressive tiers)
- Update default `maxCapacityRetries` display from 3 to 5

## Testing
- Add `tests/stress-test.cjs` for concurrent request stress testing

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

src/account-manager/index.js

@@ -15,7 +15,15 @@ import {
     markRateLimited as markLimited,
     markInvalid as markAccountInvalid,
     getMinWaitTimeMs as getMinWait,
-    getRateLimitInfo as getLimitInfo
+    getRateLimitInfo as getLimitInfo,
+    getConsecutiveFailures as getFailures,
+    resetConsecutiveFailures as resetFailures,
+    incrementConsecutiveFailures as incrementFailures,
+    markAccountCoolingDown as markCoolingDown,
+    isAccountCoolingDown as checkCoolingDown,
+    clearAccountCooldown as clearCooldown,
+    getCooldownRemaining as getCooldownMs,
+    CooldownReason
 } from './rate-limits.js';
 import {
     getTokenForAccount as fetchToken,
@@ -182,6 +190,10 @@ export class AccountManager {
         if (this.#strategy) {
             this.#strategy.onSuccess(account, modelId);
         }
+        // Reset consecutive failures on success (matches opencode-antigravity-auth)
+        if (account?.email) {
+            resetFailures(this.#accounts, account.email);
+        }
     }
 
     /**
@@ -206,6 +218,26 @@ export class AccountManager {
         }
     }
 
+    /**
+     * Get the consecutive failure count for an account
+     * Used for progressive backoff calculation
+     * @param {string} email - Account email
+     * @returns {number} Number of consecutive failures
+     */
+    getConsecutiveFailures(email) {
+        return getFailures(this.#accounts, email);
+    }
+
+    /**
+     * Increment the consecutive failure count without marking as rate limited
+     * Used for quick retries to track failures while staying on same account
+     * @param {string} email - Account email
+     * @returns {number} New consecutive failure count
+     */
+    incrementConsecutiveFailures(email) {
+        return incrementFailures(this.#accounts, email);
+    }
+
     /**
      * Get the current strategy name
      * @returns {string} Strategy name
@@ -275,6 +307,52 @@ export class AccountManager {
         return getLimitInfo(this.#accounts, email, modelId);
     }
 
+    // ============================================================================
+    // Cooldown Methods (matches opencode-antigravity-auth)
+    // ============================================================================
+
+    /**
+     * Mark an account as cooling down for a specified duration
+     * Used for temporary backoff separate from rate limits
+     * @param {string} email - Email of the account
+     * @param {number} cooldownMs - Duration of cooldown in milliseconds
+     * @param {string} [reason] - Reason for the cooldown (use CooldownReason constants)
+     */
+    markAccountCoolingDown(email, cooldownMs, reason = CooldownReason.RATE_LIMIT) {
+        markCoolingDown(this.#accounts, email, cooldownMs, reason);
+    }
+
+    /**
+     * Check if an account is currently cooling down
+     * @param {string} email - Email of the account
+     * @returns {boolean} True if account is cooling down
+     */
+    isAccountCoolingDown(email) {
+        const account = this.#accounts.find(a => a.email === email);
+        return account ? checkCoolingDown(account) : false;
+    }
+
+    /**
+     * Clear the cooldown for an account
+     * @param {string} email - Email of the account
+     */
+    clearAccountCooldown(email) {
+        const account = this.#accounts.find(a => a.email === email);
+        if (account) {
+            clearCooldown(account);
+        }
+    }
+
+    /**
+     * Get time remaining until cooldown expires for an account
+     * @param {string} email - Email of the account
+     * @returns {number} Milliseconds until cooldown expires, 0 if not cooling down
+     */
+    getCooldownRemaining(email) {
+        const account = this.#accounts.find(a => a.email === email);
+        return account ? getCooldownMs(account) : 0;
+    }
+
     /**
      * Get OAuth token for an account
      * @param {Object} account - Account object with email and credentials
@@ -378,4 +456,7 @@ export class AccountManager {
     }
 }
 
+// Re-export CooldownReason for use by handlers
+export { CooldownReason };
+
 export default AccountManager;