YazSes v2.5.0-dev.2 — Wave I ship-now tier (developer preview)¶
Pre-release · not published (publish workflows are pinned to v0.x/v1.x; v2 dev tags do not ship). v1.4.1 remains the stable release. Every feature is OFF by default.
Continues Wave I (design/vision/v2-research/11-wave-i.md) with the remaining two dependency-free ship-now cores, completing the tier:
New since dev.1 (all off by default)¶
- Structured-Markup Dictation (ADR-v2-067,
[markup]) — speak document structure and get Markdown/org markup: "bullet list: apples; oranges; pears" → a list; "table columns Name, Age; row Alice, 30" → a valid table. Distinct from Spoken Spreadsheet (drives an app). - Document Find-and-Replace (ADR-v2-068,
[findreplace]) — edit the whole document by voice, not just the last utterance: "replace every 'utilise' with 'use'", with all/first scope and a case-sensitivity flag. Impossible in a cloud tool — it has no local text access.
Plus the dev.1 cores: Terminal Command Safety Gate and Spoken Regex Builder. yazses features now lists 79 capabilities.
Wave I backlog (heavier / hardware-gated)¶
Hard Contextual Biasing (trie hotword boosting), Diarized Conversation Capture + Rename-by-Voice, Adaptive Latency Governor (speculative decoding), Per-Language Auto Model Switching, Voice Window/ Workspace Management, Citation-by-Voice from a local BibTeX/CSL library.
Quality¶
1223 tests green across Linux × macOS × Windows on Python 3.11 and 3.12. Both new cores are unit-tested to 100% with no new dependency; the default install and the v1 dictation path are unchanged.