# YazSes v2.6.0-dev.4 — Wave J continues (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 J** (`design/vision/v2-research/12-wave-j.md`) with two more pure cores:

## New since dev.3 (all off by default)

- **Voice Jump-to-Symbol** (ADR-v2-081, `[jump]`) — "jump to function tokenize", "go to line 240"
  → the editor moves there. `resolve_target` (line/symbol/search) + `fuzzy_pick` (phonetically
  fuzzy difflib match) + `plan_motion` (with a text-search fallback when no symbol matches). Fills
  a real gap — YazSes had no voice *navigation*. No camera (unlike Gaze) and no blind grid (unlike
  Mouse Grid). The live LSP symbol feed is deferred.
- **Spoken Shell Pipeline Builder** (ADR-v2-082, `[shellpipe]`) — speak stages ("list files, pipe
  to grep error, pipe to word count") → renders `ls | grep error | wc -l` **as text**; nothing
  runs until you say "run it". `parse_stages` + `render_pipeline` (curated read-only vocab +
  `shlex`-quoted grep, aborting on any unknown stage) + `dryrun_wrap`. Distinct from the Terminal
  Command Safety Gate — this *constructs* the pipeline and defaults to preview. The NL2Bash SLM is
  deferred.

`yazses features` now lists **93 capabilities**.

## Wave J remaining
Crowd-Proof Dictation (target-speaker extraction), Recording Import (batch transcription).

## Quality
1288 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.