v0.5.0

Added

• Benchmark consolidation: consolidated benchmark profile output with snapshot artifacts for repeatable analysis
• Profiling coverage: flamegraph + pyinstrument speedscope generation across all Spikard bindings
• Benchmark reporting: comparative CI benchmark summaries in README files with binding-specific breakdowns
• Runtime test clients: native WebSocket/SSE test clients for PHP and Ruby; WASM runtime WebSocket support with test coverage
• Benchmark suites: expanded FastAPI/Robyn coverage in benchmark CI and fixture-driven baselines

Changed

• Performance defaults: disabled TraceLayer/request-id/timeout/compression by default for higher baseline throughput
• Core/HTTP performance: reduced per-request parsing, skipped unused query/body work, cached schema validation, and avoided mime parsing overhead
• Binding hot paths: faster JSON and request conversion in Python/Ruby/PHP/Node/WASM, fewer allocations, and better handler fast paths
• CI focus: benchmark CI now prioritizes profiling Spikard bindings and consolidating results

Fixed

• Benchmark harness: repaired TS/Node/WASM baseline startup, multipart boundaries, Express/DTO command, and correct PID targeting for perf
• FFI safety: removed unwraps/panics across bindings and tightened error boundary handling
• E2E stability: corrected scoped package references, WASM header parsing, and Ruby gem dependency setup
• PHP e2e generator: added missing __invoke on generated handlers and normalized newline expectations in string fixtures