feat: first release with npm distribution live (v0.5.107)

- @perryts/perry and 7 per-platform optional-dep packages publish via
  OIDC Trusted Publisher from .github/workflows/release-packages.yml
  on every GitHub Release
- Covers darwin-{arm64,x64}, linux-{x64,arm64} glibc + musl, win32-x64
- `npx @perryts/perry compile file.ts` works end-to-end
- No runtime/codegen change

Author: proggeramlug

CLAUDE.md

@@ -8,7 +8,7 @@ This file provides guidance to Claude Code (claude.ai/code) when working with co
 
 Perry is a native TypeScript compiler written in Rust that compiles TypeScript source code directly to native executables. It uses SWC for TypeScript parsing and LLVM for code generation.
 
-**Current Version:** 0.5.106
+**Current Version:** 0.5.107
 
 ## TypeScript Parity Status
 
@@ -150,6 +150,7 @@ First-resolved directory cached in `compile_package_dirs`; subsequent imports re
 
 Keep entries to 1-2 lines max. Full details in CHANGELOG.md.
 
+- **v0.5.107** — First end-to-end release with npm distribution live. `@perryts/perry` + seven per-platform optional-dep packages (`@perryts/perry-{darwin-arm64,darwin-x64,linux-x64,linux-arm64,linux-x64-musl,linux-arm64-musl,win32-x64}`) publish via OIDC Trusted Publisher from `release-packages.yml` on each GitHub Release. `npx @perryts/perry compile file.ts` works on all seven platforms. No runtime/codegen change.
 - **v0.5.106** — Swap `lettre`'s `tokio1-native-tls` feature for `tokio1-rustls-tls` in `crates/perry-stdlib/Cargo.toml`. Eliminates `openssl-sys` / `native-tls` from the transitive dep tree (they were the only holdouts; the policy comment at Cargo.toml:35 already states "rustls only to avoid OpenSSL"). Unblocks the musl CI build — `openssl-sys` was failing with "Could not find openssl via pkg-config: cross-compilation unsupported" on `x86_64-unknown-linux-musl`. No functional change for SMTP clients; rustls provides the same TLS surface.
 - **v0.5.105** — `Int32Array.length` (and other typed-array `.length`) returned 0 because `js_value_length_f64` only handled NaN-boxed pointers (top16 == 0x7FFD); typed arrays sometimes flow as raw `bitcast i64 → double` with top16 == 0. Added a raw-pointer arm guarded on the Darwin mimalloc heap window (≥ 2 TB, < 128 TB) that consults `is_registered_buffer` / `lookup_typed_array_kind`.
 - **v0.5.104** — Extend the v0.5.103 inliner fix: `substitute_locals` now also walks `WeakRefNew`/`WeakRefDeref`/`FinalizationRegistryNew`/`Object{Keys,Values,Entries,FromEntries,IsFrozen,IsSealed,IsExtensible,Create}`/`ArrayFrom`/`Uint8ArrayFrom`/`IteratorToArray`/`StructuredClone`/`QueueMicrotask`/`ProcessNextTick`/`Json{Parse,Stringify}`/`ArrayIsArray`/`Math{Sqrt,Floor,Ceil,Round,Abs,Log,Log2,Log10,Log1p,Clz32,MinSpread,MaxSpread}`. Same root cause as v0.5.103 (catch-all `_ => {}` skipped these single-operand wrappers); same user-visible class of bug — inlined function bodies referenced unmapped pre-inline LocalIds and read the wrong slot. Verified `test_gap_weakref_finalization`'s `function createAndDeref() { const inner = {...}; const innerRef = new WeakRef(inner); ... }` now correctly carries `inner` through to the WeakRef after inlining.

Cargo.toml

@@ -102,7 +102,7 @@ opt-level = "s"       # Optimize for size in stdlib
 opt-level = 3
 
 [workspace.package]
-version = "0.5.106"
+version = "0.5.107"
 edition = "2021"
 license = "MIT"
 repository = "https://github.com/PerryTS/perry"