retellai-reliability-patterns

@jeremylongshore/retellai-reliability-patterns

jeremylongshore

1,004

123 forks

Updated 1/18/2026

View on GitHub

Implement Retell AI reliability patterns including circuit breakers, idempotency, and graceful degradation. Use when building fault-tolerant Retell AI integrations, implementing retry strategies, or adding resilience to production Retell AI services. Trigger with phrases like "retellai reliability", "retellai circuit breaker", "retellai idempotent", "retellai resilience", "retellai fallback", "retellai bulkhead".

Installation

$skills install @jeremylongshore/retellai-reliability-patterns

Claude Code

Cursor

Copilot

Codex

Antigravity

Details

Repositoryjeremylongshore/claude-code-plugins-plus-skills

Pathplugins/saas-packs/retellai-pack/skills/retellai-reliability-patterns/SKILL.md

Branchmain

Scoped Name@jeremylongshore/retellai-reliability-patterns

Usage

After installing, this skill will be available to your AI coding assistant.

Verify installation:

skills list

Skill Instructions

name: retellai-reliability-patterns description: | Implement Retell AI reliability patterns including circuit breakers, idempotency, and graceful degradation. Use when building fault-tolerant Retell AI integrations, implementing retry strategies, or adding resilience to production Retell AI services. Trigger with phrases like "retellai reliability", "retellai circuit breaker", "retellai idempotent", "retellai resilience", "retellai fallback", "retellai bulkhead". allowed-tools: Read, Write, Edit version: 1.0.0 license: MIT author: Jeremy Longshore jeremy@intentsolutions.io

Retell AI Reliability Patterns

Overview

Production-grade reliability patterns for Retell AI integrations.

Prerequisites

Understanding of circuit breaker pattern
opossum or similar library installed
Queue infrastructure for DLQ
Caching layer for fallbacks

Circuit Breaker

import CircuitBreaker from 'opossum';

const retellaiBreaker = new CircuitBreaker(
  async (operation: () => Promise<any>) => operation(),
  {
    timeout: 30000,
    errorThresholdPercentage: 50,
    resetTimeout: 30000,
    volumeThreshold: 10,
  }
);

// Events
retellaiBreaker.on('open', () => {
  console.warn('Retell AI circuit OPEN - requests failing fast');
  alertOps('Retell AI circuit breaker opened');
});

retellaiBreaker.on('halfOpen', () => {
  console.info('Retell AI circuit HALF-OPEN - testing recovery');
});

retellaiBreaker.on('close', () => {
  console.info('Retell AI circuit CLOSED - normal operation');
});

// Usage
async function safeRetell AICall<T>(fn: () => Promise<T>): Promise<T> {
  return retellaiBreaker.fire(fn);
}

Idempotency Keys

import { v4 as uuidv4 } from 'uuid';
import crypto from 'crypto';

// Generate deterministic idempotency key from input
function generateIdempotencyKey(
  operation: string,
  params: Record<string, any>
): string {
  const data = JSON.stringify({ operation, params });
  return crypto.createHash('sha256').update(data).digest('hex');
}

// Or use random key with storage
class IdempotencyManager {
  private store: Map<string, { key: string; expiresAt: Date }> = new Map();

  getOrCreate(operationId: string): string {
    const existing = this.store.get(operationId);
    if (existing && existing.expiresAt > new Date()) {
      return existing.key;
    }

    const key = uuidv4();
    this.store.set(operationId, {
      key,
      expiresAt: new Date(Date.now() + 24 * 60 * 60 * 1000),
    });
    return key;
  }
}

Bulkhead Pattern

import PQueue from 'p-queue';

// Separate queues for different operations
const retellaiQueues = {
  critical: new PQueue({ concurrency: 10 }),
  normal: new PQueue({ concurrency: 5 }),
  bulk: new PQueue({ concurrency: 2 }),
};

async function prioritizedRetell AICall<T>(
  priority: 'critical' | 'normal' | 'bulk',
  fn: () => Promise<T>
): Promise<T> {
  return retellaiQueues[priority].add(fn);
}

// Usage
await prioritizedRetell AICall('critical', () =>
  retellaiClient.processPayment(order)
);

await prioritizedRetell AICall('bulk', () =>
  retellaiClient.syncCatalog(products)
);

Timeout Hierarchy

const TIMEOUT_CONFIG = {
  connect: 5000,      // Initial connection
  request: 30000,     // Standard requests
  upload: 120000,     // File uploads
  longPoll: 300000,   // Webhook long-polling
};

async function timedoutRetell AICall<T>(
  operation: 'connect' | 'request' | 'upload' | 'longPoll',
  fn: () => Promise<T>
): Promise<T> {
  const timeout = TIMEOUT_CONFIG[operation];

  return Promise.race([
    fn(),
    new Promise<never>((_, reject) =>
      setTimeout(() => reject(new Error(`Retell AI ${operation} timeout`)), timeout)
    ),
  ]);
}

Graceful Degradation

interface Retell AIFallback {
  enabled: boolean;
  data: any;
  staleness: 'fresh' | 'stale' | 'very_stale';
}

async function withRetell AIFallback<T>(
  fn: () => Promise<T>,
  fallbackFn: () => Promise<T>
): Promise<{ data: T; fallback: boolean }> {
  try {
    const data = await fn();
    // Update cache for future fallback
    await updateFallbackCache(data);
    return { data, fallback: false };
  } catch (error) {
    console.warn('Retell AI failed, using fallback:', error.message);
    const data = await fallbackFn();
    return { data, fallback: true };
  }
}

Dead Letter Queue

interface DeadLetterEntry {
  id: string;
  operation: string;
  payload: any;
  error: string;
  attempts: number;
  lastAttempt: Date;
}

class Retell AIDeadLetterQueue {
  private queue: DeadLetterEntry[] = [];

  add(entry: Omit<DeadLetterEntry, 'id' | 'lastAttempt'>): void {
    this.queue.push({
      ...entry,
      id: uuidv4(),
      lastAttempt: new Date(),
    });
  }

  async processOne(): Promise<boolean> {
    const entry = this.queue.shift();
    if (!entry) return false;

    try {
      await retellaiClient[entry.operation](entry.payload);
      console.log(`DLQ: Successfully reprocessed ${entry.id}`);
      return true;
    } catch (error) {
      entry.attempts++;
      entry.lastAttempt = new Date();

      if (entry.attempts < 5) {
        this.queue.push(entry);
      } else {
        console.error(`DLQ: Giving up on ${entry.id} after 5 attempts`);
        await alertOnPermanentFailure(entry);
      }
      return false;
    }
  }
}

Health Check with Degraded State

type HealthStatus = 'healthy' | 'degraded' | 'unhealthy';

async function retellaiHealthCheck(): Promise<{
  status: HealthStatus;
  details: Record<string, any>;
}> {
  const checks = {
    api: await checkApiConnectivity(),
    circuitBreaker: retellaiBreaker.stats(),
    dlqSize: deadLetterQueue.size(),
  };

  const status: HealthStatus =
    !checks.api.connected ? 'unhealthy' :
    checks.circuitBreaker.state === 'open' ? 'degraded' :
    checks.dlqSize > 100 ? 'degraded' :
    'healthy';

  return { status, details: checks };
}

Instructions

Step 1: Implement Circuit Breaker

Wrap Retell AI calls with circuit breaker.

Step 2: Add Idempotency Keys

Generate deterministic keys for operations.

Step 3: Configure Bulkheads

Separate queues for different priorities.

Step 4: Set Up Dead Letter Queue

Handle permanent failures gracefully.

Output

Circuit breaker protecting Retell AI calls
Idempotency preventing duplicates
Bulkhead isolation implemented
DLQ for failed operations

Error Handling

Issue	Cause	Solution
Circuit stays open	Threshold too low	Adjust error percentage
Duplicate operations	Missing idempotency	Add idempotency key
Queue full	Rate too high	Increase concurrency
DLQ growing	Persistent failures	Investigate root cause

Examples

Quick Circuit Check

const state = retellaiBreaker.stats().state;
console.log('Retell AI circuit:', state);

Resources

Next Steps

For policy enforcement, see retellai-policy-guardrails.

More by jeremylongshore

View all

oauth-callback-handler

1,004

Oauth Callback Handler - Auto-activating skill for API Integration. Triggers on: oauth callback handler, oauth callback handler Part of the API Integration skill category.

rabbitmq-queue-setup

1,004

Rabbitmq Queue Setup - Auto-activating skill for Backend Development. Triggers on: rabbitmq queue setup, rabbitmq queue setup Part of the Backend Development skill category.

model-evaluation-suite

1,004

evaluating-machine-learning-models: This skill allows Claude to evaluate machine learning models using a comprehensive suite of metrics. It should be used when the user requests model performance analysis, validation, or testing. Claude can use this skill to assess model accuracy, precision, recall, F1-score, and other relevant metrics. Trigger this skill when the user mentions "evaluate model", "model performance", "testing metrics", "validation results", or requests a comprehensive "model evaluation".

neural-network-builder

1,004

building-neural-networks: This skill allows Claude to construct and configure neural network architectures using the neural-network-builder plugin. It should be used when the user requests the creation of a new neural network, modification of an existing one, or assistance with defining the layers, parameters, and training process. The skill is triggered by requests involving terms like "build a neural network," "define network architecture," "configure layers," or specific mentions of neural network types (e.g., "CNN," "RNN," "transformer").