본문으로 건너뛰기
cccskills
무료GitHub에 공개

data-service-circuit-breaker-aware

Build TypeScript data services that separate API calls from display components. Add caching and a circuit breaker, which pauses calls after repeated failures, while keeping user-triggered requests able to report errors directly.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

원문원문 설명 보기

Create data fetching services with circuit breaker pattern for API resilience. Services handle fetch, cache, retry, and expose typed data to panel components. Includes clear guidance on when to skip the circuit breaker for user-triggered one-shot calls versus background auto-polling services.

설치 방법 보기

이럴 때 유용합니다

  • Building data services for automatically refreshed panels
  • Pausing repeated calls to a failing API
  • Separating generation errors from background refresh handling
  • Keeping API keys in server-side proxy endpoints

포함된 파일(2)

  • SKILL.md8.6 KB
  • .skill_id42 B

SKILL.md(원문)

설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.

Data Service Pattern

Each panel's data comes from a dedicated service module in src/services/. Services handle API calls, caching, error handling with circuit breaker pattern.

Circuit Breaker

The circuit breaker prevents cascading failures when an API is down. After N consecutive failures, it enters "cooldown" mode and serves cached data instead of hitting the API.

Create src/utils/circuit-breaker.ts:

interface CircuitState {
  failures: number;
  cooldownUntil: number;
}

interface CacheEntry<T> {
  data: T;
  timestamp: number;
}

export interface CircuitBreakerOptions {
  name: string;
  maxFailures?: number;
  cooldownMs?: number;
  cacheTtlMs?: number;
}

export class CircuitBreaker<T> {
  private state: CircuitState = { failures: 0, cooldownUntil: 0 };
  private cache: CacheEntry<T> | null = null;
  private name: string;
  private maxFailures: number;
  private cooldownMs: number;
  private cacheTtlMs: number;

  constructor(options: CircuitBreakerOptions) {
    this.name = options.name;
    this.maxFailures = options.maxFailures ?? 2;
    this.cooldownMs = options.cooldownMs ?? 5 * 60 * 1000;  // 5 minutes
    this.cacheTtlMs = options.cacheTtlMs ?? 10 * 60 * 1000; // 10 minutes
  }

  isOnCooldown(): boolean {
    if (Date.now() < this.state.cooldownUntil) return true;
    if (this.state.cooldownUntil > 0) {
      this.state = { failures: 0, cooldownUntil: 0 };
    }
    return false;
  }

  getCached(): T | null {
    if (this.cache && Date.now() - this.cache.timestamp < this.cacheTtlMs) {
      return this.cache.data;
    }
    return null;
  }

  recordSuccess(data: T): void {
    this.state = { failures: 0, cooldownUntil: 0 };
    this.cache = { data, timestamp: Date.now() };
  }

  recordFailure(error?: string): void {
    this.state.failures++;
    if (this.state.failures >= this.maxFailures) {
      this.state.cooldownUntil = Date.now() + this.cooldownMs;
      console.warn(`[${this.name}] Cooldown for ${this.cooldownMs / 1000}s`);
    }
  }

  async execute<R extends T>(fn: () => Promise<R>, defaultValue: R): Promise<R> {
    if (this.isOnCooldown()) {
      const cached = this.getCached();
      return (cached as R) ?? defaultValue;
    }
    const cached = this.getCached();
    if (cached !== null) return cached as R;
    try {
      const result = await fn();
      this.recordSuccess(result);
      return result;
    } catch (e) {
      console.error(`[${this.name}] Failed:`, e);
      this.recordFailure(String(e));
      return defaultValue;
    }
  }
}

export function createCircuitBreaker<T>(options: CircuitBreakerOptions): CircuitBreaker<T> {
  return new CircuitBreaker<T>(options);
}

Service Module Pattern

Each service is a TypeScript module that exports async data-fetching functions:

// src/services/stock-market.ts
import { createCircuitBreaker } from '../utils/circuit-breaker';

export interface StockQuote {
  symbol: string;
  name: string;
  price: number | null;
  change: number | null;
  sparkline?: number[];
}

const breaker = createCircuitBreaker<StockQuote[]>({
  name: 'StockMarket',
  cacheTtlMs: 60_000, // 1 minute cache
});

export async function fetchStockQuotes(symbols: string[]): Promise<StockQuote[]> {
  return breaker.execute(async () => {
    const params = new URLSearchParams({ symbols: symbols.join(',') });
    const resp = await fetch(`/api/stocks?${params}`);
    if (!resp.ok) throw new Error(`HTTP ${resp.status}`);
    return resp.json();
  }, []);
}

Key Patterns

  1. One circuit breaker per API endpoint — create at module level
  2. Export typed interfaces for data shapes
  3. Wrap fetch calls in breaker.execute(fn, defaultValue)
  4. Default values are empty arrays [] or empty objects {}
  5. Services are stateless modules — no class instances, just exported functions
  6. API proxy: Frontend calls /api/... which proxies to external APIs (hides API keys)

When to Skip the Circuit Breaker

The circuit breaker is designed to protect background auto-polling services from cascading failures. It is not always appropriate — applying it blindly adds unnecessary complexity and can actively cause problems for user-driven flows.

Use the circuit breaker when:

  • The service is polled automatically on an interval (e.g., every 30 seconds, every minute)
  • Repeated failures would hammer a degraded API endpoint
  • Serving stale cached data is an acceptable fallback (the user won't notice a brief outage)
  • The data refreshes in the background without explicit user intent
// ✅ Background polling — circuit breaker appropriate
const breaker = createCircuitBreaker<StockQuote[]>({ name: 'StockMarket', cacheTtlMs: 60_000 });

export async function fetchStockQuotes(symbols: string[]): Promise<StockQuote[]> {
  return breaker.execute(async () => {
    const resp = await fetch(`/api/stocks?symbols=${symbols.join(',')}`);
    if (!resp.ok) throw new Error(`HTTP ${resp.status}`);
    return resp.json();
  }, []);
}

// Called on a timer — circuit breaker protects against rapid failure storms
setInterval(() => fetchStockQuotes(['AAPL', 'MSFT']), 60_000);

Skip the circuit breaker when:

  • The call is user-triggered (button click, explicit refresh, form submit)
  • The operation is one-shot and non-repeating (e.g., LLM text generation, send message)
  • Failure should surface immediately to the user as an error — not silently return stale data
  • The user expects a fresh result, not a cache hit
  • Retry/error handling is managed by the UI layer (e.g., a "Try again" button)
// ✅ User-triggered one-shot LLM call — NO circuit breaker
export interface DailyBriefingData {
  briefing: string;
  generatedAt: string;
}

export async function generateDailyBriefing(context: string): Promise<DailyBriefingData> {
  // No circuit breaker: this is explicitly triggered by the user clicking "Generate".
  // Failures should throw so the UI can show an error and offer a retry button.
  const resp = await fetch('/api/llm/generate', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({ prompt: context }),
  });
  if (!resp.ok) throw new Error(`LLM generation failed: HTTP ${resp.status}`);
  const { text } = await resp.json();
  return { briefing: text, generatedAt: new Date().toISOString() };
}

Decision table

TriggerRepeats automatically?Stale data OK?Use circuit breaker?
Timer / interval✅ Yes✅ Yes✅ Yes
App startup fetch❌ No✅ YesOptional
Button / user action❌ No❌ No❌ No
LLM generation❌ No❌ No❌ No
Explicit refresh❌ No❌ No❌ No

UI error handling for one-shot calls

When skipping the circuit breaker, the calling component is responsible for surfacing errors:

// In the panel/component that calls a one-shot service:
async generate(): Promise<void> {
  if (this.isFetching) return;
  this.isFetching = true;
  this.showLoading();
  try {
    const data = await generateDailyBriefing(this.buildContext());
    this.render(data);
  } catch (e) {
    // Surface the error directly — no silent cache fallback
    this.showError('Generation failed. Please try again.', () => this.generate());
  } finally {
    this.isFetching = false;
  }
}

Rule of thumb: If a human is waiting for the result of a specific action they just took, skip the circuit breaker and let errors propagate to the UI. If a machine is polling in the background, use the circuit breaker.

API Proxy Pattern

For APIs requiring keys, create server-side proxy endpoints:

// api/stocks.ts (Vercel serverless function)
export default async function handler(req, res) {
  const { symbols } = req.query;
  const apiKey = process.env.FINNHUB_API_KEY;
  const data = await fetch(`https://finnhub.io/api/v1/quote?symbol=${symbols}&token=${apiKey}`);
  res.json(await data.json());
}

This keeps API keys server-side and provides a single endpoint for the frontend.

리뷰

아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.

댓글

로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.

아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.

보내면 바로 공개됩니다. 이메일 주소나 전화번호 같은 개인 정보는 쓰지 마세요. URL은 2개까지입니다.

0 / 1,000자

이 양식은 reCAPTCHA로 보호되며 Google 개인정보처리방침 및 서비스 약관이 적용됩니다.

같은 저장소의 스킬

Prepare audio stems, the separate parts of a track, for mixing by matching their durations with silence, loops, or crossfades. Verify each generated and processed file before moving to the next production step.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Match audio parts with different durations
  • Extend drums and melodic parts using suitable methods
  • Check audio specifications before mixing
HKUDS/OpenSpace7.8천2026년 8월 13일 업데이트

Produce audio in separate stems, meaning individual instrument tracks, and verify each output as you go. Align track lengths with silence, loops, or crossfades, then apply effects, mix a master, and package the files.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Align instrument tracks to a reference audio file
  • Choose how to extend tracks that are too short
  • Verify audio properties after generation and effects
HKUDS/OpenSpace7.8천2026년 8월 13일 업데이트

Build serverless endpoints that relay external API requests for a dashboard while keeping API keys on the server. Includes Vercel examples for input checks, caching, cross-origin access, and consistent JSON responses.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Connect a dashboard to external APIs without exposing API keys
  • Create endpoints for stock quotes and news
  • Standardize API responses and error handling
HKUDS/OpenSpace7.8천2026년 8월 13일 업데이트

Produce a master track and separate audio parts through a structured Python workflow. Apply filters and reverb, export WAV files with consistent settings, package them in a ZIP archive, and check that the deliverables meet the requested specifications.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Export a master track with separate audio parts
  • Apply filters and reverb to audio
  • Check duration and format across deliverables
HKUDS/OpenSpace7.8천2026년 8월 13일 업데이트

Keep document creation moving when web searches and page retrieval repeatedly fail. The skill preserves the original task, uses the model’s existing knowledge, and clearly identifies information that could not be checked against external sources.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Continue drafting after repeated web retrieval failures
  • Preserve task requirements during network errors
  • Mark information that could not be externally verified
HKUDS/OpenSpace7.8천2026년 8월 13일 업데이트

When Python execution repeatedly fails in execute_code_sandbox, use a file-based alternative: save the script, run it through the shell, and capture results or errors. The guidance covers paths, imports, timeouts, structured output, and optional cleanup.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

  • Retry Python work through the shell after repeated sandbox failures
  • Run longer tasks with an adjusted timeout
  • Execute Python work that needs multiple files or system utilities
HKUDS/OpenSpace7.8천2026년 8월 13일 업데이트

HKUDS의 스킬 모두 보기

이 스킬의 문제 신고