johannes/assistant
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(vela): start mocked response flow after push-to-talk commit

Browse Source
This commit is contained in:
Johannes Kresner
2026-04-08 21:20:17 +02:00
parent 98bcc543f5
commit 28712443cc
8 changed files with 284 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
docs/backlog.md
View File

@@ -190,9 +190,9 @@ Polish the system after the core voice loop is reliable.
- `apps/vela-gateway` now exposes a minimal `/ws` WebSocket session skeleton with ephemeral in-memory sessions and defensive message handling
- `apps/vela-gateway` now accepts `mocked.turn.trigger` and emits protocol-valid mocked transcript/response events with one in-flight mocked turn per session
- `apps/vela-gateway` now supports placeholder input-audio append/commit cycles before running another mocked turn on the same socket
- `apps/vela-gateway` now emits deterministic `transcript.partial` events for placeholder `input_audio.append` messages and exactly one deterministic `transcript.final` for each placeholder `input_audio.commit`
- `apps/vela-ui` now renders the latest placeholder partial transcript during the push-to-talk shell turn and replaces it with the deterministic final transcript on commit
- `apps/vela-ui` now exposes a cancel control for active mocked turns and keeps already-rendered transcript/response text visible after cancellation
- `apps/vela-gateway` now honors `response.cancel` during mocked turns by stopping pending mocked response events, returning the session to `idle`, and allowing a new mocked turn on the same socket
- `apps/vela-gateway` now emits deterministic `transcript.partial` events for placeholder `input_audio.append` messages and, after each accepted `input_audio.commit`, reuses the mocked response engine to stream a deterministic assistant reply for that push-to-talk turn
- `apps/vela-ui` now renders the latest placeholder partial transcript during the push-to-talk shell turn, replaces it with the deterministic final transcript on commit, and shows streamed assistant text for the same push-to-talk flow
- `apps/vela-ui` now exposes a cancel control for active mocked turns and mocked push-to-talk responses, and keeps already-rendered transcript/response text visible after cancellation
- `apps/vela-gateway` now honors `response.cancel` during mocked turns and push-to-talk-triggered mocked responses by stopping pending mocked response events, returning the session to `idle`, and allowing a new turn on the same socket
- `apps/vela-protocol` now provides the shared WebSocket event contract for the UI and gateway
- backend framework choice is now concrete: Fastify