Copilot API Integration Skill

Description

## Copilot API Integration Skill

### Overview
Use Copilot to generate robust API client code with proper error handling, retry logic, and type-safe request/response handling.

### Method 1: From API Documentation
Describe the API endpoint and let Copilot generate the client:

```typescript
// Generate a typed API client for the Stripe Payments API
// POST /v1/charges
// Headers: Authorization: Bearer sk_test_xxx, Content-Type: application/x-www-form-urlencoded
// Body: amount (int, cents), currency (string), source (string)
// Response: Charge object with id, amount, status, created

async function createCharge(params: CreateChargeParams): Promise<Charge> {
  // Copilot generates the complete implementation
}
```

### Method 2: SDK Wrapper Generation
Generate a complete API client class:

```
Generate a TypeScript API client class for a REST API with:
- Base URL configuration
- JWT authentication with automatic token refresh
- Request/response interceptors for logging
- Retry logic with exponential backoff (3 retries)
- Timeout configuration (30s default)
- Type-safe methods for: GET, POST, PUT, PATCH, DELETE
- Error handling that wraps HTTP errors in typed exceptions
```

### Method 3: From OpenAPI/Swagger Spec
```
@workspace Generate a TypeScript API client from the OpenAPI spec
at docs/openapi.yaml. Include:
- Interface for each schema definition
- Method for each endpoint with proper typing
- Request and response types
- JSDoc comments from API descriptions
```

### Robust Error Handling Pattern
```typescript
// Generate an API client with comprehensive error handling:
// - Network errors: retry with exponential backoff
// - 401: refresh token and retry once
// - 429: respect Retry-After header
// - 4xx: throw typed validation/not-found errors
// - 5xx: retry up to 3 times, then throw
// - Timeout: throw TimeoutError after 30s
```

### Rate Limiting and Throttling
```typescript
// Generate a rate-limited API client that:
// - Respects X-RateLimit-Remaining and X-RateLimit-Reset headers
// - Queues requests when approaching rate limit
// - Automatically backs off when rate limited
// - Supports concurrent request limits (max 10 in-flight)
```

### Webhook Handler Generation
```typescript
// Generate a webhook handler for Stripe events:
// - Verify webhook signature using signing secret
// - Parse event type and route to appropriate handler
// - Idempotency: skip already-processed events
// - Error handling: return 200 even on processing failure
// - Logging: log event type, id, and processing result
```

### Testing API Integrations
```
Generate tests for the PaymentClient:
- Mock HTTP responses for success and error cases
- Test retry logic with sequential failure then success
- Test token refresh on 401 response
- Test timeout handling
- Test rate limit handling with Retry-After header
- Use MSW (Mock Service Worker) for realistic HTTP mocking
```

### Tips
- Always generate types for request/response payloads
- Include retry logic for production API clients
- Handle all HTTP status codes, not just 200
- Add request/response logging for debugging
- Use dependency injection for the HTTP client to enable testing

Free to copy and use. Compatible with GitHub Copilot, GPT-5.

How to Use This Skill

Tags

Compatible Agents

Version History

Install Command

Related Prompts