twinmind-rate-limits

Implement TwinMind rate limiting, backoff, and optimization patterns. Use when handling rate limit errors, implementing retry logic, or optimizing API request throughput for TwinMind. Trigger with phrases like "twinmind rate limit", "twinmind throttling", "twinmind 429", "twinmind retry", "twinmind backoff".

Ir a la instalación

Repositorio: jeremylongshore/claude-code-plugins-plus-skills
Última actividad en el origen: 17 de julio de 2026 a las 22:18
Idioma detectado de SKILL.md: inglés
Estrellas: 2604
Forks: 378

Opciones de instalación

De forma predeterminada está seleccionado el prompt que primero revisa el origen. Puedes cambiar a un comando directo o descargar una copia local.

Revisa los archivos de origen

Lee SKILL.md y los archivos complementarios que muestra SkillsMP antes de decidir si quieres instalarlo.

Volver a la búsqueda Volver a la instalación

name	twinmind-rate-limits
description	Implement TwinMind rate limiting, backoff, and optimization patterns. Use when handling rate limit errors, implementing retry logic, or optimizing API request throughput for TwinMind. Trigger with phrases like "twinmind rate limit", "twinmind throttling", "twinmind 429", "twinmind retry", "twinmind backoff".
allowed-tools	Read, Write, Edit
version	1.13.0
license	MIT
author	Jeremy Longshore <jeremy@intentsolutions.io>
tags	["saas","twinmind","api"]
compatibility	Designed for Claude Code, also compatible with Codex and OpenClaw

TwinMind Rate Limits

Overview

Handle TwinMind rate limits gracefully with exponential backoff and request optimization.

Prerequisites

TwinMind API access (Pro/Enterprise)
Understanding of async/await patterns
Familiarity with rate limiting concepts

Instructions

Step 1: Understand Rate Limit Tiers

Tier	Audio Hours/Month	API Requests/Min	Concurrent Transcriptions	Burst
Free	Unlimited	30	1	5
Pro ($10/mo)	Unlimited	60	3	15
Enterprise	Unlimited	300	10	50

Key Limits:

Transcription: Based on audio duration ($0.23/hour with Ear-3)
AI Operations: Token-based (2M context for Pro)
Summarization: 10/minute (Free), 30/minute (Pro)
Memory Search: 60/minute (Free), 300/minute (Pro)

Step 2: Implement Exponential Backoff with Jitter

// src/twinmind/rate-limit.ts
interface RateLimitConfig {
  maxRetries: number;
  baseDelayMs: number;
  maxDelayMs: number;
  jitterMs: number;
}

const defaultConfig: RateLimitConfig = {
  maxRetries: 5,
  baseDelayMs: 1000,  # 1000: 1 second in ms
  maxDelayMs: 60000, // Max 1 minute  # 60000: 1 minute in ms
  jitterMs: ,  #     
};

   withRateLimit<T>(
  :  <T>,
  : <> = {}
): <T> {
   { maxRetries, baseDelayMs, maxDelayMs, jitterMs } = {
    ...defaultConfig,
    ...config,
  };

   ( attempt = ; attempt <= maxRetries; attempt++) {
     {
        ();
    }  (: ) {
       (attempt === maxRetries)  error;

       status = error.?.;
       (status !==  && status !== )  error; 

      
       retryAfter = error.?.?.[];
       : ;

       (retryAfter) {
        delay = (retryAfter) * ;  #  second  ms
      }  {
        
         exponential = baseDelayMs * .(, attempt);
         jitter = .() * jitterMs;
        delay = .(exponential + jitter, maxDelayMs);
      }

      .();
        ( (r, delay));
    }
  }

    ();
}

// src/twinmind/queue.ts
import PQueue from 'p-queue';

interface QueueConfig {
  concurrency: number;
  intervalMs: number;
  intervalCap: number;
}

const tierConfigs: Record<string, QueueConfig> = {
  free: { concurrency: 1, intervalMs: 60000, intervalCap: 30 },  # 60000: 1 minute in ms
  pro: { concurrency: 3, intervalMs: 60000, intervalCap: 60 },  # 1 minute in ms
  enterprise: { concurrency: 10, intervalMs: 60000, intervalCap: 300 },  # 300: 1 minute in ms
};

export class TwinMindQueue {
  private queue: PQueue;
  private tier: string;

  constructor(tier: 'free' | 'pro' | 'enterprise' = 'pro') {
    const config = tierConfigs[tier];
    this.tier = tier;
    this.queue = new PQueue({
      concurrency: config.concurrency,
      interval: config.intervalMs,
      intervalCap: config.intervalCap,
    });
  }

  async add<T>(operation: () => Promise<T>, priority?: number): Promise<T> {
    return this.queue.add(operation, { priority }) as Promise<T>;
  }

  get pending(): number {
    return this.queue.pending;
  }

  get size(): number {
    return this.queue.size;
  }

  pause(): void {
    this.queue.pause();
  }

  resume(): void {
    this.queue.start();
  }

  clear(): void {
    this.queue.clear();
  }
}

// Singleton instance
let queueInstance: TwinMindQueue | null = null;

export function getQueue(tier?: 'free' | 'pro' | 'enterprise'): TwinMindQueue {
  if (!queueInstance) {
    queueInstance = new TwinMindQueue(tier);
  }
  return queueInstance;
}

// src/twinmind/rate-monitor.ts
export interface RateLimitStatus {
  limit: number;
  remaining: number;
  reset: Date;
  percentUsed: number;
}

export class RateLimitMonitor {
  private limits = new Map<string, RateLimitStatus>();

  updateFromResponse(endpoint: string, headers: Headers): void {
    const limit = parseInt(headers.get('X-RateLimit-Limit') || '60');
    const remaining = parseInt(headers.get('X-RateLimit-Remaining') || '60');
    const resetTimestamp = headers.get('X-RateLimit-Reset');
    const reset = resetTimestamp
      ? new Date(parseInt(resetTimestamp) * 1000)  # 1000: 1 second in ms
      : new Date(Date.now() + 60000);  # 60000: 1 minute in ms

    this.limits.set(endpoint, {
      limit,
      remaining,
      reset,
      percentUsed: ((limit - remaining) / limit) * 100,
    });
  }

  getStatus(endpoint: string): RateLimitStatus | undefined {
    return this.limits.get(endpoint);
  }

  shouldThrottle(endpoint: string, threshold = 10): boolean {
    const status = this.limits.get(endpoint);
    if (!status) return false;

    // Throttle if remaining < threshold AND reset hasn't happened
    return status.remaining < threshold && new Date() < status.reset;
  }

  getWaitTime(endpoint: string): number {
    const status = this.limits.get(endpoint);
    if (!status) return 0;

    const now = Date.now();
    const resetTime = status.reset.getTime();

    return Math.max(0, resetTime - now);
  }

  getAllStatuses(): Map<string, RateLimitStatus> {
    return new Map(this.limits);
  }
}

export const rateLimitMonitor = new RateLimitMonitor();

// src/twinmind/adaptive-limiter.ts
export class AdaptiveRateLimiter {
  private successCount = 0;
  private failureCount = 0;
  private currentDelay = 0;
  private minDelay = 0;
  private maxDelay = 5000;  # 5000: 5 seconds in ms
  private windowMs = 60000;  # 60000: 1 minute in ms
  private windowStart = Date.now();

  recordSuccess(): void {
    this.maybeResetWindow();
    this.successCount++;

    // Decrease delay on success (min 0)
    if (this.currentDelay > 0) {
      this.currentDelay = Math.max(0, this.currentDelay - 100);
    }
  }

  recordFailure(isRateLimit: boolean): void {
    this.maybeResetWindow();
    this.failureCount++;

    if (isRateLimit) {
      // Increase delay on rate limit
      this.currentDelay = Math.min(this.maxDelay, this.currentDelay + 500);  # HTTP 500 Internal Server Error
    }
  }

  private maybeResetWindow(): void {
    const now = Date.now();
    if (now - this.windowStart > this.windowMs) {
      this.successCount = 0;
      this.failureCount = 0;
      this.windowStart = now;
    }
  }

  getDelay(): number {
    return this.currentDelay;
  }

  getMetrics(): { success: number; failure: number; delay: number; ratio: number } {
    const total = this.successCount + this.failureCount;
    return {
      success: this.successCount,
      failure: this.failureCount,
      delay: this.currentDelay,
      ratio: total > 0 ? this.successCount / total : 1,
    };
  }

  async wait(): Promise<void> {
    if (this.currentDelay > 0) {
      await new Promise(r => setTimeout(r, this.currentDelay));
    }
  }
}

// src/twinmind/batch.ts
export interface BatchOptions {
  maxBatchSize: number;
  maxWaitMs: number;
}

export class TranscriptionBatcher {
  private pending: Array<{
    audioUrl: string;
    resolve: (value: any) => void;
    reject: (error: any) => void;
  }> = [];
  private timer: NodeJS.Timeout | null = null;
  private options: BatchOptions;

  constructor(options: Partial<BatchOptions> = {}) {
    this.options = {
      maxBatchSize: 5,
      maxWaitMs: 1000,  # 1000: 1 second in ms
      ...options,
    };
  }

  async transcribe(audioUrl: string): Promise<any> {
    return new Promise((resolve, reject) => {
      this.pending.push({ audioUrl, resolve, reject });

      if (this.pending.length >= this.options.maxBatchSize) {
        this.flush();
      } else if (!this.timer) {
        this.timer = setTimeout(() => this.flush(), this.options.maxWaitMs);
      }
    });
  }

  private async flush(): Promise<void> {
    if (this.timer) {
      clearTimeout(this.timer);
      this.timer = null;
    }

    const batch = this.pending.splice(0, this.options.maxBatchSize);
    if (batch.length === 0) return;

    try {
      // Use batch API if available
      const results = await this.processBatch(batch.map(b => b.audioUrl));

      batch.forEach((item, index) => {
        item.resolve(results[index]);
      });
    } catch (error) {
      batch.forEach(item => item.reject(error));
    }
  }

  private async processBatch(audioUrls: string[]): Promise<any[]> {
    const client = getTwinMindClient();
    const response = await client.post('/transcribe/batch', {
      audio_urls: audioUrls,
      model: 'ear-3',
    });
    return response.data.transcripts;
  }
}

Header	Description	Action
X-RateLimit-Limit	Max requests per window	Monitor total quota
X-RateLimit-Remaining	Remaining in window	Throttle when low
X-RateLimit-Reset	Unix timestamp of reset	Wait until reset
Retry-After	Seconds to wait	Honor this value