* feat(pi-codex-conversion): add native Codex voice * refactor(pi-codex-conversion): refine realtime voice UI * feat(pi-codex-conversion): add configurable voice sessions * feat(pi-codex-conversion): add portable voice configuration * fix(pi-codex-conversion): simplify voice setup copy * feat(pi-codex-conversion): guide voice device setup * feat(pi-codex-conversion): include realtime prompt in setup * fix(pi-codex-conversion): simplify guided voice setup * feat(pi-codex-conversion): layer workspace voice prompts * docs(pi-codex-conversion): clarify voice prompt boundary * feat(pi-codex-conversion): add voice shortcuts * refactor(pi-codex-conversion): harden voice ownership * fix(pi-codex-conversion): close voice session gaps * fix(ci): provision voice helper builds * fix(ci): rebuild binaries after workflow changes * Bundle Codex tool binaries * fix(pi-codex-conversion): launch web search executable on Windows * fix(pi-codex-conversion): close voice startup gaps * fix(pi-codex-conversion): harden voice startup portability * chore(pi-codex-conversion): ship runtime-only Rust assets * perf(pi-codex-conversion): defer voice transport loading * perf(pi-codex-conversion): defer usage client loading * chore: update Pi dependencies to 0.82.0 * fix(pi-codex-conversion): close voice provider gaps * Bundle Codex tool binaries * fix(pi-codex-conversion): recover lost websocket continuations * fix(pi-codex-conversion): clarify interruptible commands * fix(pi-codex-conversion): stop voice immediately * fix(pi-codex-conversion): bound voice helper output * refactor(pi-codex-conversion): cull performative voice tests * fix(pi-codex-conversion): stop dictation on trigger release * fix(pi-codex-conversion): infer dictation key release --------- Co-authored-by: Test <test@example.com> Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>