attio-performance-tuning

# Attio Performance Tuning ## Overview Attio's REST API returns JSON over HTTPS. Performance optimization focuses on reducing request count (batching, caching), maximizing throughput (connection reuse, pagination), and minimizing latency (selective field fetching, parallel queries). ## Key Performance Facts | Factor | Detail | |--------|--------| | Rate limit | Sliding 10-second window, shared across all tokens | | Pagination default | `limit: 500` (max per page) | | API base | `https://api.attio.com/v2` | | Auth overhead | Bearer token in header (minimal) | | Response format | JSON only (no binary/protobuf) | ## Instructions ### Strategy 1: Response Caching with LRU Cache read-heavy data (object schemas, attribute definitions) that rarely change: ```typescript import { LRUCache } from "lru-cache"; const cache = new LRUCache<string, unknown>({ max: 500, // Max entries ttl: 5 * 60 * 1000, // 5 minutes for schema data }); async function cachedGet<T>( client: AttioClient, path: string, ttlMs?: number ): Promise<T> { const cached = cache.get(path) as T | undefined; if (cached) return cached; const result = await client.get<T>(path); cache.set(path, result, { ttl: ttlMs }); return result; } // Schema data: cache for 30 minutes (rarely changes) const objects = await cachedGet(client, "/objects", 30 * 60 * 1000); const attrs = await cachedGet(client, "/objects/people/attributes", 30 * 60 * 1000); // Record data: cache for 1-5 minutes (...