attio-rate-limits

# Attio Rate Limits ## Overview Attio uses a **sliding window algorithm** with a **10-second window**. Rate limit scores are summed across all apps and access tokens hitting the API. When exceeded, you get HTTP 429 with a `Retry-After` header containing a date (usually the next second). Attio may temporarily reduce limits during incidents. ## Rate Limit Response ``` HTTP/1.1 429 Too Many Requests Retry-After: Sat, 22 Mar 2025 14:30:01 GMT Content-Type: application/json { "status_code": 429, "type": "rate_limit_error", "code": "rate_limit_exceeded", "message": "Rate limit exceeded, please try again later" } ``` **Key fact:** The `Retry-After` header is a date string (not seconds). Parse it as a Date to calculate wait time. ## Instructions ### Step 1: Parse Retry-After Header ```typescript function parseRetryAfter(headers: Headers): number { const retryAfter = headers.get("Retry-After"); if (!retryAfter) return 1000; // Default 1s // Attio sends a date string const retryDate = new Date(retryAfter); const waitMs = retryDate.getTime() - Date.now(); return Math.max(waitMs, 100); // Minimum 100ms } ``` ### Step 2: Exponential Backoff with Retry-After Awareness ```typescript import { AttioApiError } from "./client"; interface RetryConfig { maxRetries: number; baseMs: number; maxMs: number; } async function withRateLimitRetry<T>( operation: () => Promise<{ data: T; headers?: Headers }>, config: RetryConfig = { maxRetries: 5, baseMs: 1000, m...