Release v1.1.0

• Optimized buffer allocation in the memory-mapped I/O module, reducing peak memory overhead by 15% during heavy serialization tasks.
• Resolved a race condition in the asynchronous task runner that occasionally caused thread deadlocks during multi-threaded batch processing.
• Added support for custom header injection in the transport layer API, enabling granular control over request-response metadata.