# Specifications

Implementation-ready designs: what gets built, and how it is judged done.

!!! info "This is the engineering tier"

    These pages are written for contributors, reviewers and researchers rather
    than for people using YazSes. For the user documentation, start at
    [Home](../../index.md).

**11 documents.**

- [Spec: Cocktail Filter — Target-Speaker Gate & Suppression](cocktail-filter.md)
- [Spec: Diarized Recording Import — `yazses transcribe <file>`](diarized-recording-import.md)
- [Spec: Ghost Ahead — Endpoint / Turn Anticipation (latency hiding)](ghost-ahead.md)
- [Spec: Glance-Type — Coarse Gaze-Targeted Dictation](glance-type.md)
- [Spec: Mid-Thought Undo — Spoken Buffer Revision](mid-thought-undo.md)
- [Spec: Polyglot Switch — Configured-Pair Code-Switching ASR](polyglot-switch.md)
- [Spec: Prosody Ink — prosody-driven text formatting](prosody-ink.md)
- [Spec: Punch-In — re-speak-to-correct span splicing](punch-in.md)
- [Spec: Read-Back Loop — eyes-free dictation with offline TTS confirm/correct](read-back-loop.md)
- [Spec: Say-Macro — User-Programmable Voice Launcher](say-macro.md)
- [Spec: Voiceprint Mind — On-Device Personal Adaptation](voiceprint-mind.md)