# Changelog
All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [0.1.1] - 2026-05-12
### Added
- Pool size management: `max_pool_size` and `min_pool_size` with alert callback
- `workers()` to list all connected workers with full info
- `drain_worker()` for graceful rolling deploys (no new tasks, finish existing)
- `remove_worker()` to force-disconnect and fail pending tasks
- `dispatch_to()` for targeted dispatch to a specific worker
- `dispatch_with_tag()` for tag-based routing
- Worker tags/metadata support in registration
- `PoolError::WorkerAtCapacity` and `PoolError::WorkerUnavailable` error variants
### Fixed
- Heartbeat no longer overwrites Draining status
- Re-registration of existing workers allowed even at max pool capacity
- Min-pool callback only fires on worker loss, not during initial ramp-up
- `dispatch_to()` propagates specific error (at capacity vs draining vs not found)
## [0.1.0] - 2026-05-11
### Added
- Core dispatcher with builder pattern API
- Worker pool with least-loaded selection algorithm
- Task schema with priority, timeout, retry, and trace ID support
- WebSocket transport for networked workers
- Unix stdio transport for CLI tool integration
- Docker transport for containerized workers
- SSH transport for remote workers
- WASM transport for sandboxed execution
- Input validation to prevent command injection
- Dead worker detection via heartbeat timeout
- Backpressure signaling from workers
- camelCase JSON wire protocol for cross-language compatibility
- SDKs: TypeScript, Python, Go, Java, C#, Ruby, Elixir, Swift, PHP
- 185 tests with 95.58% code coverage