The SDK shifted toward tighter integration with Scrapy's async primitives and more robust request handling. Version 3.3.0 migrated to Scrapy's native AsyncCrawlerRunner and introduced Actor schema storage aliases, while subsequent patches addressed cascading failure modes in cleanup, fixed Scrapy request round-trip conversion, and eliminated race conditions in request queue operations. The team also hardened configuration validation to stop treating zero as falsy and prevented deadlocks when the API fails during request state transitions.