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

feat(vela): retire legacy mocked turn trigger

Browse Source
This commit is contained in:
Johannes Kresner
2026-04-08 21:50:18 +02:00
parent 28712443cc
commit 8e14eaeed0
10 changed files with 78 additions and 378 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
apps/vela-gateway/src/index.js
View File

@@ -11,7 +11,6 @@ const {
const WEBSOCKET_ROUTE = '/ws'; const WEBSOCKET_ROUTE = '/ws';
const WEBSOCKET_GUID = '258EAFA5-E914-47DA-95CA-C5AB0DC85B11'; const WEBSOCKET_GUID = '258EAFA5-E914-47DA-95CA-C5AB0DC85B11';
const MOCKED_USER_TRANSCRIPT = '[mocked user] What is the current mocked vertical slice?';
const MOCKED_ASSISTANT_RESPONSE = '[mocked assistant] This is a deterministic mocked response from the gateway vertical slice.'; const MOCKED_ASSISTANT_RESPONSE = '[mocked assistant] This is a deterministic mocked response from the gateway vertical slice.';
function createPlaceholderPartialTranscript(audioChunkCount) { function createPlaceholderPartialTranscript(audioChunkCount) {
@@ -67,7 +66,7 @@ function scheduleMockedTurnStep(session, turnId, delay, callback) {
session.mockedTurnTimers.push(timer); session.mockedTurnTimers.push(timer);
} }
function startMockedTurn(socket, session, { transcript = MOCKED_USER_TRANSCRIPT, includeListeningState = true } = {}) { function startMockedTurn(socket, session, { transcript, includeListeningState = true } = {}) {
if (session.mockedTurnInFlight) { if (session.mockedTurnInFlight) {
sendSocketError(socket, 'mocked_turn_in_flight', 'Only one mocked turn can run per session at a time.'); sendSocketError(socket, 'mocked_turn_in_flight', 'Only one mocked turn can run per session at a time.');
return; return;
@@ -246,7 +245,11 @@ function handleClientMessage(socket, session, rawMessage) {
sendSocketMessage(socket, 'session.state', { value: session.state }); sendSocketMessage(socket, 'session.state', { value: session.state });
break; break;
case 'mocked.turn.trigger': case 'mocked.turn.trigger':
startMockedTurn(socket, session); sendSocketError(
socket,
'unsupported_mocked_turn_trigger',
'mocked.turn.trigger is no longer supported; use input_audio.append and input_audio.commit instead.'
);
break; break;
case 'input_audio.append': case 'input_audio.append':
if (session.mockedTurnInFlight) { if (session.mockedTurnInFlight) {

121
apps/vela-gateway/test/websocket-session.test.js
View File

@@ -355,7 +355,7 @@ test('websocket handles valid and invalid client messages safely', async () => {
} }
}); });
test('websocket accepts a placeholder input cycle before a mocked turn on the same socket', async () => { test('websocket accepts repeated placeholder input cycles on the same socket', async () => {
const server = await startServer(); const server = await startServer();
try { try {
@@ -403,11 +403,15 @@ test('websocket accepts a placeholder input cycle before a mocked turn on the sa
payload: { value: 'idle' } payload: { value: 'idle' }
}); });
client.sendJson({ type: 'mocked.turn.trigger', payload: {} }); client.sendJson({ type: 'input_audio.append', payload: { chunk: 'placeholder-control-shell-chunk-2' } });
assert.deepEqual(await client.nextMessage(), { assert.deepEqual(await client.nextMessage(), {
type: 'session.state', type: 'session.state',
payload: { value: 'listening' } payload: { value: 'listening' }
}); });
assert.deepEqual(await client.nextMessage(), {
type: 'transcript.partial',
payload: { text: '[mocked partial] Placeholder push-to-talk transcript in progress.' }
});
await client.close(); await client.close();
} finally { } finally {
@@ -516,7 +520,7 @@ test('websocket emits deterministic partials for repeated appends and a determin
} }
}); });
test('websocket mocked turn emits deterministic transcript and response events in order', async () => { test('websocket rejects the retired mocked.turn.trigger path deterministically', async () => {
const server = await startServer(); const server = await startServer();
try { try {
@@ -526,110 +530,17 @@ test('websocket mocked turn emits deterministic transcript and response events i
client.sendJson({ type: 'mocked.turn.trigger', payload: {} }); client.sendJson({ type: 'mocked.turn.trigger', payload: {} });
assert.deepEqual(await client.nextMessage(), {
type: 'session.state',
payload: { value: 'listening' }
});
assert.deepEqual(await client.nextMessage(), {
type: 'transcript.final',
payload: { text: '[mocked user] What is the current mocked vertical slice?' }
});
assert.deepEqual(await client.nextMessage(), {
type: 'session.state',
payload: { value: 'thinking' }
});
assert.deepEqual(await client.nextMessage(), {
type: 'session.state',
payload: { value: 'speaking' }
});
assert.deepEqual(await client.nextMessage(), {
type: 'response.text.delta',
payload: { text: '[mocked assistant] ' }
});
assert.deepEqual(await client.nextMessage(), {
type: 'response.text.delta',
payload: { text: 'This is a deterministic mocked response from the gateway vertical slice.' }
});
assert.deepEqual(await client.nextMessage(), {
type: 'response.completed',
payload: {}
});
assert.deepEqual(await client.nextMessage(), {
type: 'session.state',
payload: { value: 'idle' }
});
await client.close();
} finally {
await server.close();
}
});
test('websocket rejects a second mocked turn while one is in flight', async () => {
const server = await startServer();
try {
const client = await connectWebSocket(server.port);
await client.nextMessage();
await client.nextMessage();
client.sendJson({ type: 'mocked.turn.trigger', payload: {} });
assert.deepEqual(await client.nextMessage(), {
type: 'session.state',
payload: { value: 'listening' }
});
client.sendJson({ type: 'mocked.turn.trigger', payload: {} });
assert.deepEqual(await client.nextMessage(), { assert.deepEqual(await client.nextMessage(), {
type: 'error', type: 'error',
payload: { payload: {
code: 'mocked_turn_in_flight', code: 'unsupported_mocked_turn_trigger',
message: 'Only one mocked turn can run per session at a time.', message:
'mocked.turn.trigger is no longer supported; use input_audio.append and input_audio.commit instead.',
retryable: true retryable: true
} }
}); });
await client.close();
} finally {
await server.close();
}
});
test('websocket cancel stops an active mocked turn and allows a new one without reconnecting', async () => {
const server = await startServer();
try {
const client = await connectWebSocket(server.port);
await client.nextMessage();
await client.nextMessage();
client.sendJson({ type: 'mocked.turn.trigger', payload: {} });
assert.deepEqual(await client.nextMessage(), {
type: 'session.state',
payload: { value: 'listening' }
});
assert.deepEqual(await client.nextMessage(), {
type: 'transcript.final',
payload: { text: '[mocked user] What is the current mocked vertical slice?' }
});
assert.deepEqual(await client.nextMessage(), {
type: 'session.state',
payload: { value: 'thinking' }
});
client.sendJson({ type: 'response.cancel', payload: {} });
assert.deepEqual(await client.nextMessage(), {
type: 'session.state',
payload: { value: 'idle' }
});
await assert.rejects(() => client.nextMessage(150), /timed out waiting for websocket message/); await assert.rejects(() => client.nextMessage(150), /timed out waiting for websocket message/);
client.sendJson({ type: 'mocked.turn.trigger', payload: {} });
assert.deepEqual(await client.nextMessage(), {
type: 'session.state',
payload: { value: 'listening' }
});
await client.close(); await client.close();
} finally { } finally {
await server.close(); await server.close();
@@ -671,11 +582,15 @@ test('websocket cancel stops a push-to-talk commit response and allows another t
}); });
await assert.rejects(() => client.nextMessage(150), /timed out waiting for websocket message/); await assert.rejects(() => client.nextMessage(150), /timed out waiting for websocket message/);
client.sendJson({ type: 'mocked.turn.trigger', payload: {} }); client.sendJson({ type: 'input_audio.append', payload: { chunk: 'chunk-2' } });
assert.deepEqual(await client.nextMessage(), { assert.deepEqual(await client.nextMessage(), {
type: 'session.state', type: 'session.state',
payload: { value: 'listening' } payload: { value: 'listening' }
}); });
assert.deepEqual(await client.nextMessage(), {
type: 'transcript.partial',
payload: { text: '[mocked partial] Placeholder push-to-talk transcript in progress.' }
});
await client.close(); await client.close();
} finally { } finally {
@@ -694,11 +609,15 @@ test('websocket safely accepts cancel when no turn is active', async () => {
client.sendJson({ type: 'response.cancel', payload: {} }); client.sendJson({ type: 'response.cancel', payload: {} });
await assert.rejects(() => client.nextMessage(150), /timed out waiting for websocket message/); await assert.rejects(() => client.nextMessage(150), /timed out waiting for websocket message/);
client.sendJson({ type: 'mocked.turn.trigger', payload: {} }); client.sendJson({ type: 'input_audio.append', payload: { chunk: 'chunk-1' } });
assert.deepEqual(await client.nextMessage(), { assert.deepEqual(await client.nextMessage(), {
type: 'session.state', type: 'session.state',
payload: { value: 'listening' } payload: { value: 'listening' }
}); });
assert.deepEqual(await client.nextMessage(), {
type: 'transcript.partial',
payload: { text: '[mocked partial] Placeholder push-to-talk transcript in progress.' }
});
await client.close(); await client.close();
} finally { } finally {

2
apps/vela-ui/README.md
View File

@@ -7,7 +7,7 @@ Current status:
- SvelteKit app boots in the Yarn workspace - SvelteKit app boots in the Yarn workspace
- root page shows a minimal voice-session shell with connect/disconnect controls - root page shows a minimal voice-session shell with connect/disconnect controls
- the shell can connect to the gateway `/ws` endpoint and display developer-visible session status - the shell can connect to the gateway `/ws` endpoint and display developer-visible session status
- the shell can trigger one deterministic mocked turn and render the mocked transcript plus assistant response - the current mocked interaction path is push-to-talk only and renders the mocked transcript plus assistant response
- Vitest covers connect/disconnect plus the deterministic mocked transcript/response UI flow without requiring a browser harness - Vitest covers connect/disconnect plus the deterministic mocked transcript/response UI flow without requiring a browser harness
- Playwright remains optional for deeper browser-level checks - Playwright remains optional for deeper browser-level checks
- microphone capture and audio playback remain future increments - microphone capture and audio playback remain future increments

95
apps/vela-ui/e2e/voice-session.spec.js
View File

@@ -1,87 +1,15 @@
import { expect, test } from '@playwright/test'; import { expect, test } from '@playwright/test';
const MOCKED_USER_TRANSCRIPT = '[mocked user] What is the current mocked vertical slice?';
const MOCKED_ASSISTANT_RESPONSE = const MOCKED_ASSISTANT_RESPONSE =
'[mocked assistant] This is a deterministic mocked response from the gateway vertical slice.'; '[mocked assistant] This is a deterministic mocked response from the gateway vertical slice.';
test('voice session shell covers the mocked transcript/response slice', async ({ page }) => { test('voice session shell supports the placeholder push-to-talk path', async ({ page }) => {
await page.goto('/'); await page.goto('/');
await expect(page.getByTestId('hydration-status')).toHaveText('ready'); await expect(page.getByTestId('hydration-status')).toHaveText('ready');
await expect(page.getByTestId('connection-state')).toHaveText('not connected'); await expect(page.getByTestId('connection-state')).toHaveText('not connected');
await expect(page.getByTestId('mocked-turn-button')).toBeDisabled();
await expect(page.getByTestId('session-id')).toHaveText('not assigned');
await expect(page.getByTestId('gateway-session-state')).toHaveText('not received');
await page.getByTestId('connect-button').click();
await expect(page.getByTestId('connection-state')).toHaveText('connected');
await expect(page.getByTestId('gateway-session-state')).toHaveText('idle');
await expect(page.getByTestId('session-id')).not.toHaveText('not assigned');
await expect(page.getByTestId('mocked-turn-button')).toBeEnabled();
const sessionId = await page.getByTestId('session-id').textContent();
await page.getByTestId('mocked-turn-button').click();
await expect(page.getByTestId('mocked-turn-status')).toHaveText('running');
await expect(page.getByTestId('user-transcript')).toHaveText('waiting for mocked transcript…');
await expect(page.getByTestId('assistant-response')).toHaveText('waiting for mocked response…');
await expect(page.getByTestId('user-transcript')).toHaveText(MOCKED_USER_TRANSCRIPT);
await expect(page.getByTestId('assistant-response')).toHaveText(MOCKED_ASSISTANT_RESPONSE);
await expect(page.getByTestId('conversation-render-order')).toHaveText('transcript>response');
await expect(page.getByTestId('mocked-turn-status')).toHaveText('idle');
await page.getByTestId('disconnect-button').click();
await expect(page.getByTestId('connection-state')).toHaveText('disconnected');
await expect(page.getByTestId('connection-detail')).toHaveText('Gateway WebSocket is closed.');
await expect(page.getByTestId('gateway-session-state')).toHaveText('idle');
await expect(page.getByTestId('session-id')).toHaveText(sessionId ?? '');
await expect(page.getByTestId('mocked-turn-button')).toBeDisabled();
await expect(page.getByTestId('user-transcript')).toHaveText(MOCKED_USER_TRANSCRIPT);
await expect(page.getByTestId('assistant-response')).toHaveText(MOCKED_ASSISTANT_RESPONSE);
await expect(page.getByTestId('session-id')).toHaveText(sessionId ?? '');
await expect(page.getByTestId('gateway-session-state')).toHaveText('idle');
await expect(page.getByTestId('user-transcript')).toHaveText(MOCKED_USER_TRANSCRIPT);
await expect(page.getByTestId('assistant-response')).toHaveText(MOCKED_ASSISTANT_RESPONSE);
});
test('voice session shell can cancel an active mocked turn and start another one', async ({ page }) => {
await page.goto('/');
await expect(page.getByTestId('hydration-status')).toHaveText('ready');
await expect(page.getByTestId('cancel-turn-button')).toBeDisabled();
await page.getByTestId('connect-button').click();
await expect(page.getByTestId('connection-state')).toHaveText('connected');
await expect(page.getByTestId('mocked-turn-button')).toBeEnabled();
await page.getByTestId('mocked-turn-button').click();
await expect(page.getByTestId('mocked-turn-status')).toHaveText('running');
await expect(page.getByTestId('cancel-turn-button')).toBeEnabled();
await expect(page.getByTestId('user-transcript')).toHaveText(MOCKED_USER_TRANSCRIPT);
await expect(page.getByTestId('assistant-response')).toContainText('[mocked assistant]');
await page.getByTestId('cancel-turn-button').click();
await expect(page.getByTestId('gateway-session-state')).toHaveText('idle');
await expect(page.getByTestId('mocked-turn-status')).toHaveText('idle');
await expect(page.getByTestId('cancel-turn-button')).toBeDisabled();
await expect(page.getByTestId('mocked-turn-button')).toBeEnabled();
await expect(page.getByTestId('user-transcript')).toHaveText(MOCKED_USER_TRANSCRIPT);
await expect(page.getByTestId('assistant-response')).toContainText('[mocked assistant]');
await page.getByTestId('mocked-turn-button').click();
await expect(page.getByTestId('mocked-turn-status')).toHaveText('running');
await expect(page.getByTestId('assistant-response')).toHaveText(MOCKED_ASSISTANT_RESPONSE);
await expect(page.getByTestId('mocked-turn-status')).toHaveText('idle');
});
test('voice session shell supports a placeholder mic-control cycle before another mocked turn', async ({ page }) => {
await page.goto('/');
await expect(page.getByTestId('hydration-status')).toHaveText('ready');
await expect(page.getByTestId('mic-control-button')).toBeDisabled(); await expect(page.getByTestId('mic-control-button')).toBeDisabled();
await expect(page.getByTestId('mocked-turn-button')).toHaveCount(0);
await page.getByTestId('connect-button').click(); await page.getByTestId('connect-button').click();
@@ -95,10 +23,6 @@ test('voice session shell supports a placeholder mic-control cycle before anothe
await expect(page.getByTestId('partial-transcript')).toHaveText( await expect(page.getByTestId('partial-transcript')).toHaveText(
'[mocked partial] Placeholder push-to-talk transcript in progress.' '[mocked partial] Placeholder push-to-talk transcript in progress.'
); );
await expect(page.getByTestId('user-transcript')).toHaveText(
'[mocked partial] Placeholder push-to-talk transcript in progress.'
);
await expect(page.getByTestId('mocked-turn-button')).toBeDisabled();
await page.getByTestId('mic-control-button').dispatchEvent('mouseup'); await page.getByTestId('mic-control-button').dispatchEvent('mouseup');
await expect(page.getByTestId('mic-control-status')).toHaveText('idle'); await expect(page.getByTestId('mic-control-status')).toHaveText('idle');
@@ -108,15 +32,10 @@ test('voice session shell supports a placeholder mic-control cycle before anothe
await expect(page.getByTestId('assistant-response')).toHaveText(MOCKED_ASSISTANT_RESPONSE); await expect(page.getByTestId('assistant-response')).toHaveText(MOCKED_ASSISTANT_RESPONSE);
await expect(page.getByTestId('partial-transcript')).toHaveText('none'); await expect(page.getByTestId('partial-transcript')).toHaveText('none');
await expect(page.getByTestId('gateway-session-state')).toHaveText('idle'); await expect(page.getByTestId('gateway-session-state')).toHaveText('idle');
await expect(page.getByTestId('mocked-turn-button')).toBeEnabled(); await expect(page.getByTestId('mic-control-button')).toBeEnabled();
await page.getByTestId('mocked-turn-button').click();
await expect(page.getByTestId('mocked-turn-status')).toHaveText('running');
await expect(page.getByTestId('assistant-response')).toHaveText(MOCKED_ASSISTANT_RESPONSE);
await expect(page.getByTestId('mocked-turn-status')).toHaveText('idle');
}); });
test('voice session shell can cancel a push-to-talk mocked response and start another turn', async ({ page }) => { test('voice session shell can cancel a push-to-talk mocked response and start another push-to-talk turn', async ({ page }) => {
await page.goto('/'); await page.goto('/');
await expect(page.getByTestId('hydration-status')).toHaveText('ready'); await expect(page.getByTestId('hydration-status')).toHaveText('ready');
@@ -140,9 +59,11 @@ test('voice session shell can cancel a push-to-talk mocked response and start an
'[mocked final] Placeholder push-to-talk transcript completed from 1 appended chunk.' '[mocked final] Placeholder push-to-talk transcript completed from 1 appended chunk.'
); );
await expect(page.getByTestId('assistant-response')).toContainText('[mocked assistant]'); await expect(page.getByTestId('assistant-response')).toContainText('[mocked assistant]');
await expect(page.getByTestId('mocked-turn-button')).toBeEnabled(); await expect(page.getByTestId('mic-control-button')).toBeEnabled();
await page.getByTestId('mocked-turn-button').click(); await page.getByTestId('mic-control-button').dispatchEvent('mousedown');
await expect(page.getByTestId('gateway-session-state')).toHaveText('listening');
await page.getByTestId('mic-control-button').dispatchEvent('mouseup');
await expect(page.getByTestId('assistant-response')).toHaveText(MOCKED_ASSISTANT_RESPONSE); await expect(page.getByTestId('assistant-response')).toHaveText(MOCKED_ASSISTANT_RESPONSE);
await expect(page.getByTestId('mocked-turn-status')).toHaveText('idle'); await expect(page.getByTestId('mocked-turn-status')).toHaveText('idle');
}); });

44
apps/vela-ui/src/lib/VoiceSessionShell.svelte
View File

@@ -58,15 +58,6 @@
let micControlActive = false; let micControlActive = false;
let hydrationStatus = 'mounting'; let hydrationStatus = 'mounting';
$: canTriggerMockedTurn =
typeof WebSocket !== 'undefined' &&
connectionState === 'connected' &&
socket?.readyState === WebSocket.OPEN &&
sessionReadyReceived &&
gatewaySessionState === 'idle' &&
!micControlActive &&
!mockedTurnInFlight;
$: canCancelMockedTurn = $: canCancelMockedTurn =
typeof WebSocket !== 'undefined' && typeof WebSocket !== 'undefined' &&
connectionState === 'connected' && connectionState === 'connected' &&
@@ -132,33 +123,6 @@
socket.send(JSON.stringify(createMessageEnvelope('input_audio.commit', {}))); socket.send(JSON.stringify(createMessageEnvelope('input_audio.commit', {})));
} }
function triggerMockedTurn() {
if (!socket || socket.readyState !== WebSocket.OPEN || connectionState !== 'connected') {
connectionDetail = 'Connect to the gateway before triggering a mocked turn.';
lastError = 'mocked turn requires an active WebSocket connection';
return;
}
if (!sessionReadyReceived) {
connectionDetail = 'Wait for the gateway session to be ready before triggering a mocked turn.';
lastError = 'mocked turn requires session.ready';
return;
}
if (mockedTurnInFlight) {
connectionDetail = 'A mocked turn is already running for this session.';
return;
}
mockedUserTranscript = 'waiting for mocked transcript…';
inProgressPartialTranscript = 'none';
mockedAssistantResponse = 'waiting for mocked response…';
mockedTurnInFlight = true;
assistantResponseExpected = true;
lastError = 'none';
socket.send(JSON.stringify(createMessageEnvelope('mocked.turn.trigger', {})));
}
function cancelActiveResponse() { function cancelActiveResponse() {
if (!socket || socket.readyState !== WebSocket.OPEN || connectionState !== 'connected') { if (!socket || socket.readyState !== WebSocket.OPEN || connectionState !== 'connected') {
connectionDetail = 'Connect to the gateway before cancelling a mocked turn.'; connectionDetail = 'Connect to the gateway before cancelling a mocked turn.';
@@ -363,9 +327,8 @@
<p class="eyebrow">Vela UI</p> <p class="eyebrow">Vela UI</p>
<h1>Voice session shell</h1> <h1>Voice session shell</h1>
<p> <p>
This minimal browser shell can connect to the gateway WebSocket, trigger one deterministic This minimal browser shell can connect to the gateway WebSocket and expose a push-to-talk
mocked turn, and expose a push-to-talk control shell that only sends placeholder protocol control shell that only sends placeholder protocol events for the active session.
events for the active session.
</p> </p>
<p class="contract-note"> <p class="contract-note">
@@ -408,9 +371,6 @@
> >
Disconnect Disconnect
</button> </button>
<button data-testid="mocked-turn-button" on:click={triggerMockedTurn} disabled={!canTriggerMockedTurn}>
Run mocked turn
</button>
<button data-testid="cancel-turn-button" on:click={cancelActiveResponse} disabled={!canCancelMockedTurn}> <button data-testid="cancel-turn-button" on:click={cancelActiveResponse} disabled={!canCancelMockedTurn}>
Cancel active turn Cancel active turn
</button> </button>

121
apps/vela-ui/tests/voice-session.test.js
View File

@@ -67,7 +67,7 @@ describe('voice session shell', () => {
render(VoiceSessionShell); render(VoiceSessionShell);
expect(getByTestId('connection-state').textContent).toBe('not connected'); expect(getByTestId('connection-state').textContent).toBe('not connected');
expect(getByTestId('mocked-turn-button').hasAttribute('disabled')).toBe(true); expect(screen.queryByTestId('mocked-turn-button')).toBeNull();
expect(getByTestId('mic-control-button').hasAttribute('disabled')).toBe(true); expect(getByTestId('mic-control-button').hasAttribute('disabled')).toBe(true);
await fireEvent.click(getByTestId('connect-button')); await fireEvent.click(getByTestId('connect-button'));
@@ -78,7 +78,7 @@ describe('voice session shell', () => {
await waitFor(() => { await waitFor(() => {
expect(getByTestId('connection-state').textContent).toBe('connected'); expect(getByTestId('connection-state').textContent).toBe('connected');
expect(getByTestId('mocked-turn-button').hasAttribute('disabled')).toBe(false); expect(screen.queryByTestId('mocked-turn-button')).toBeNull();
expect(getByTestId('mic-control-button').hasAttribute('disabled')).toBe(false); expect(getByTestId('mic-control-button').hasAttribute('disabled')).toBe(false);
}); });
@@ -87,13 +87,13 @@ describe('voice session shell', () => {
await waitFor(() => { await waitFor(() => {
expect(getByTestId('connection-state').textContent).toBe('disconnected'); expect(getByTestId('connection-state').textContent).toBe('disconnected');
expect(getByTestId('mocked-turn-button').hasAttribute('disabled')).toBe(true); expect(screen.queryByTestId('mocked-turn-button')).toBeNull();
expect(getByTestId('mic-control-button').hasAttribute('disabled')).toBe(true); expect(getByTestId('mic-control-button').hasAttribute('disabled')).toBe(true);
expect(getByTestId('session-id').textContent).toBe('session-123'); expect(getByTestId('session-id').textContent).toBe('session-123');
}); });
}); });
it('runs a placeholder mic-control cycle, streams mocked assistant text, and keeps mocked turn usable on the same socket', async () => { it('runs a placeholder mic-control cycle, streams mocked assistant text, and leaves push-to-talk usable on the same socket', async () => {
render(VoiceSessionShell); render(VoiceSessionShell);
await fireEvent.click(getByTestId('connect-button')); await fireEvent.click(getByTestId('connect-button'));
@@ -131,7 +131,6 @@ describe('voice session shell', () => {
await waitFor(() => { await waitFor(() => {
expect(getByTestId('gateway-session-state').textContent).toBe('listening'); expect(getByTestId('gateway-session-state').textContent).toBe('listening');
expect(getByTestId('mocked-turn-button').hasAttribute('disabled')).toBe(true);
expect(getByTestId('user-transcript').textContent).toBe( expect(getByTestId('user-transcript').textContent).toBe(
'[mocked partial] Placeholder push-to-talk transcript in progress.' '[mocked partial] Placeholder push-to-talk transcript in progress.'
); );
@@ -168,7 +167,7 @@ describe('voice session shell', () => {
await waitFor(() => { await waitFor(() => {
expect(getByTestId('gateway-session-state').textContent).toBe('idle'); expect(getByTestId('gateway-session-state').textContent).toBe('idle');
expect(getByTestId('mocked-turn-button').hasAttribute('disabled')).toBe(false); expect(getByTestId('mic-control-button').hasAttribute('disabled')).toBe(false);
expect(getByTestId('user-transcript').textContent).toBe( expect(getByTestId('user-transcript').textContent).toBe(
'[mocked final] Placeholder push-to-talk transcript completed from 1 appended chunk.' '[mocked final] Placeholder push-to-talk transcript completed from 1 appended chunk.'
); );
@@ -177,11 +176,6 @@ describe('voice session shell', () => {
); );
expect(getByTestId('partial-transcript').textContent).toBe('none'); expect(getByTestId('partial-transcript').textContent).toBe('none');
}); });
await fireEvent.click(getByTestId('mocked-turn-button'));
expect(socket.sent).toHaveLength(3);
expect(JSON.parse(socket.sent[2]).type).toBe('mocked.turn.trigger');
}); });
it('keeps rendered push-to-talk transcript and assistant text visible after cancel', async () => { it('keeps rendered push-to-talk transcript and assistant text visible after cancel', async () => {
@@ -239,118 +233,33 @@ describe('voice session shell', () => {
}); });
}); });
it('renders mocked transcript before assistant response for a connected session', async () => { it('shows protocol errors returned for the retired mocked turn trigger path', async () => {
render(VoiceSessionShell); render(VoiceSessionShell);
await fireEvent.click(getByTestId('connect-button')); await fireEvent.click(getByTestId('connect-button'));
const socket = MockWebSocket.latest(); const socket = MockWebSocket.latest();
socket.open(); socket.open();
socket.message(createMessageEnvelope('session.ready', { sessionId: 'session-456' })); socket.message(createMessageEnvelope('session.ready', { sessionId: 'session-legacy' }));
socket.message(createMessageEnvelope('session.state', { value: 'idle' })); socket.message(createMessageEnvelope('session.state', { value: 'idle' }));
await waitFor(() => { await waitFor(() => {
expect(getByTestId('connection-state').textContent).toBe('connected'); expect(getByTestId('connection-state').textContent).toBe('connected');
}); });
await fireEvent.click(getByTestId('mocked-turn-button'));
expect(socket.sent).toHaveLength(1);
const sentMessage = JSON.parse(socket.sent[0]);
expect(sentMessage.type).toBe('mocked.turn.trigger');
socket.message(createMessageEnvelope('session.state', { value: 'listening' }));
socket.message( socket.message(
createMessageEnvelope('transcript.partial', { createMessageEnvelope('error', {
text: '[mocked partial] Placeholder push-to-talk transcript in progress.' code: 'unsupported_mocked_turn_trigger',
message:
'mocked.turn.trigger is no longer supported; use input_audio.append and input_audio.commit instead.',
retryable: true
}) })
); );
socket.message(createMessageEnvelope('transcript.final', { text: 'Turn on the office lamp.' }));
socket.message(createMessageEnvelope('session.state', { value: 'thinking' }));
socket.message(createMessageEnvelope('session.state', { value: 'speaking' }));
socket.message(createMessageEnvelope('response.text.delta', { text: 'Mocked ' }));
socket.message(createMessageEnvelope('response.text.delta', { text: 'assistant response.' }));
socket.message(createMessageEnvelope('response.completed', { reason: 'mocked_turn_complete' }));
socket.message(createMessageEnvelope('session.state', { value: 'idle' }));
await waitFor(() => { await waitFor(() => {
expect(getByTestId('user-transcript').textContent).toBe('Turn on the office lamp.'); expect(getByTestId('last-error').textContent).toBe(
expect(getByTestId('partial-transcript').textContent).toBe('none'); 'unsupported_mocked_turn_trigger: mocked.turn.trigger is no longer supported; use input_audio.append and input_audio.commit instead.'
expect(getByTestId('assistant-response').textContent).toBe('Mocked assistant response.'); );
expect(getByTestId('conversation-render-order').textContent).toBe('transcript>response');
expect(getByTestId('gateway-session-state').textContent).toBe('idle'); expect(getByTestId('gateway-session-state').textContent).toBe('idle');
}); });
}); });
it('blocks mocked turn trigger before session.ready and allows it after session.ready', async () => {
render(VoiceSessionShell);
await fireEvent.click(getByTestId('connect-button'));
const socket = MockWebSocket.latest();
socket.open();
await waitFor(() => {
expect(getByTestId('connection-state').textContent).toBe('connected');
});
expect(getByTestId('mocked-turn-button').hasAttribute('disabled')).toBe(true);
await fireEvent.click(getByTestId('mocked-turn-button'));
expect(socket.sent).toHaveLength(0);
expect(getByTestId('last-error').textContent).toBe('mocked turn requires session.ready');
socket.message(createMessageEnvelope('session.ready', { sessionId: 'session-789' }));
socket.message(createMessageEnvelope('session.state', { value: 'idle' }));
await waitFor(() => {
expect(getByTestId('mocked-turn-button').hasAttribute('disabled')).toBe(false);
});
await fireEvent.click(getByTestId('mocked-turn-button'));
expect(socket.sent).toHaveLength(1);
expect(JSON.parse(socket.sent[0]).type).toBe('mocked.turn.trigger');
});
it('shows cancel control during an active mocked turn and preserves rendered text after cancel', async () => {
render(VoiceSessionShell);
await fireEvent.click(getByTestId('connect-button'));
const socket = MockWebSocket.latest();
socket.open();
socket.message(createMessageEnvelope('session.ready', { sessionId: 'session-cancel' }));
socket.message(createMessageEnvelope('session.state', { value: 'idle' }));
await waitFor(() => {
expect(getByTestId('cancel-turn-button').hasAttribute('disabled')).toBe(true);
});
await fireEvent.click(getByTestId('mocked-turn-button'));
expect(JSON.parse(socket.sent[0]).type).toBe('mocked.turn.trigger');
socket.message(createMessageEnvelope('session.state', { value: 'listening' }));
socket.message(createMessageEnvelope('transcript.final', { text: 'Keep this transcript.' }));
socket.message(createMessageEnvelope('session.state', { value: 'thinking' }));
socket.message(createMessageEnvelope('session.state', { value: 'speaking' }));
socket.message(createMessageEnvelope('response.text.delta', { text: 'Partial response' }));
await waitFor(() => {
expect(getByTestId('cancel-turn-button').hasAttribute('disabled')).toBe(false);
expect(getByTestId('assistant-response').textContent).toBe('Partial response');
});
await fireEvent.click(getByTestId('cancel-turn-button'));
expect(JSON.parse(socket.sent[1]).type).toBe('response.cancel');
socket.message(createMessageEnvelope('session.state', { value: 'idle' }));
await waitFor(() => {
expect(getByTestId('mocked-turn-status').textContent).toBe('idle');
expect(getByTestId('gateway-session-state').textContent).toBe('idle');
expect(getByTestId('cancel-turn-button').hasAttribute('disabled')).toBe(true);
expect(getByTestId('mocked-turn-button').hasAttribute('disabled')).toBe(false);
expect(getByTestId('user-transcript').textContent).toBe('Keep this transcript.');
expect(getByTestId('assistant-response').textContent).toBe('Partial response');
});
});
}); });

10
docs/architecture.md
View File

@@ -36,14 +36,13 @@ The repository now includes separate runnable workspaces for the UI and gateway
- PWA enabled - PWA enabled
- WebSocket client - WebSocket client
The current implementation is a minimal SvelteKit app with a single voice-session shell page. The shipped UI can open and close a browser WebSocket connection to the gateway `/ws` endpoint, show explicit connection status (`not connected`, `connecting`, `connected`, `disconnected`, `error`), expose mic control shell interactions that emit placeholder `input_audio.append` / `input_audio.commit` events, trigger one deterministic mocked turn while connected, render deterministic placeholder partial/final transcripts for the push-to-talk shell, and stream the mocked assistant response both for `mocked.turn.trigger` and for push-to-talk commits. This remains a shell only: there is no real microphone capture, real provider integration, or audio playback yet. The current implementation is a minimal SvelteKit app with a single voice-session shell page. The shipped UI can open and close a browser WebSocket connection to the gateway `/ws` endpoint, show explicit connection status (`not connected`, `connecting`, `connected`, `disconnected`, `error`), expose mic control shell interactions that emit placeholder `input_audio.append` / `input_audio.commit` events, render deterministic placeholder partial/final transcripts for the push-to-talk shell, and stream the mocked assistant response after push-to-talk commit. This remains a shell only: there is no real microphone capture, real provider integration, or audio playback yet.
#### Responsibilities #### Responsibilities
Current shell responsibilities: Current shell responsibilities:
- connection state rendering - connection state rendering
- mocked-turn trigger rendering with disconnected/in-flight guards
- mocked transcript and mocked assistant response rendering - mocked transcript and mocked assistant response rendering
- developer-oriented session metadata rendering - developer-oriented session metadata rendering
- browser session connect/disconnect controls - browser session connect/disconnect controls
@@ -62,7 +61,6 @@ Current shell:
- developer-focused voice-session panel - developer-focused voice-session panel
- connect button - connect button
- disconnect button - disconnect button
- mocked-turn button
- connection status indicator - connection status indicator
- mocked transcript display - mocked transcript display
- mocked assistant response display - mocked assistant response display
@@ -106,14 +104,14 @@ The current implementation is a minimal Fastify service with `/`, `/health`, and
- the gateway sends `session.ready` followed by `session.state` (`idle`) when the socket is established - the gateway sends `session.ready` followed by `session.state` (`idle`) when the socket is established
- valid minimal client events, including placeholder `input_audio.append` / `input_audio.commit`, can move the session through the mocked turn states on one socket - valid minimal client events, including placeholder `input_audio.append` / `input_audio.commit`, can move the session through the mocked turn states on one socket
- placeholder `input_audio.append` emits deterministic mocked `transcript.partial` events and `input_audio.commit` emits one deterministic mocked `transcript.final` before starting the existing mocked assistant response stream - placeholder `input_audio.append` emits deterministic mocked `transcript.partial` events and `input_audio.commit` emits one deterministic mocked `transcript.final` before starting the existing mocked assistant response stream
- `mocked.turn.trigger` drives a fixed transcript/response event sequence over the existing shared protocol
- only one mocked turn is allowed in flight per session at a time - only one mocked turn is allowed in flight per session at a time
- invalid JSON, invalid envelopes, and malformed frames are handled defensively so the process stays up - invalid JSON, invalid envelopes, and malformed frames are handled defensively so the process stays up
- retired `mocked.turn.trigger` messages are rejected with a deterministic recoverable error
### Current UI shell behavior ### Current UI shell behavior
- renders a minimal developer-focused voice-session panel - renders a minimal developer-focused voice-session panel
- exposes connect, disconnect, mic-control shell interactions, and mocked-turn controls - exposes connect, disconnect, and mic-control shell interactions
- does not request microphone permission or capture real microphone audio - does not request microphone permission or capture real microphone audio
- only emits placeholder `input_audio.append` / `input_audio.commit` events; it does not send real audio data or play back audio - only emits placeholder `input_audio.append` / `input_audio.commit` events; it does not send real audio data or play back audio
- renders the latest placeholder partial transcript during a push-to-talk shell turn, replaces it with the final deterministic transcript on commit, and appends streamed mocked assistant text for that same push-to-talk turn - renders the latest placeholder partial transcript during a push-to-talk shell turn, replaces it with the final deterministic transcript on commit, and appends streamed mocked assistant text for that same push-to-talk turn
@@ -122,7 +120,7 @@ The current implementation is a minimal Fastify service with `/`, `/health`, and
## Voice Pipeline ## Voice Pipeline
```text ```text
Mic control shell / mocked turn button → Placeholder `input_audio.append` / `input_audio.commit` or mocked session flow → Deterministic transcript events → Shared mocked response engine → Mocked response text events → UI Mic control shell → Placeholder `input_audio.append` / `input_audio.commit` → Deterministic transcript events → Shared mocked response engine → Mocked response text events → UI
``` ```
This mocked vertical slice intentionally stands in for the future real pipeline: This mocked vertical slice intentionally stands in for the future real pipeline:

13
docs/backlog.md
View File

@@ -183,16 +183,15 @@ Polish the system after the core voice loop is reliable.
- `apps/vela-ui` now boots as a minimal SvelteKit app with a starter page - `apps/vela-ui` now boots as a minimal SvelteKit app with a starter page
- `apps/vela-ui` now includes a minimal voice-session shell that can connect to the gateway `/ws` endpoint and display developer-visible session status - `apps/vela-ui` now includes a minimal voice-session shell that can connect to the gateway `/ws` endpoint and display developer-visible session status
- `apps/vela-ui` can now trigger one deterministic mocked turn while connected and render the mocked transcript plus assistant response for the active session
- `apps/vela-ui` now exposes a visible push-to-talk mic control shell that sends placeholder `input_audio.append` / `input_audio.commit` events without requesting browser mic permission or capturing real audio - `apps/vela-ui` now exposes a visible push-to-talk mic control shell that sends placeholder `input_audio.append` / `input_audio.commit` events without requesting browser mic permission or capturing real audio
- `apps/vela-ui` now includes browser-level coverage for the mocked transcript/response slice, including connect, disconnect, and disconnected-state trigger guarding - `apps/vela-ui` now includes browser-level coverage for the placeholder push-to-talk mocked transcript/response slice, including connect, disconnect, and cancel behavior
- `apps/vela-gateway` now boots as a minimal Fastify app with `/` and `/health` endpoints - `apps/vela-gateway` now boots as a minimal Fastify app with `/` and `/health` endpoints
- `apps/vela-gateway` now exposes a minimal `/ws` WebSocket session skeleton with ephemeral in-memory sessions and defensive message handling - `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 rejects retired `mocked.turn.trigger` requests with a deterministic recoverable error instead of starting a mocked turn
- `apps/vela-gateway` now supports placeholder input-audio append/commit cycles before running another mocked turn on the same socket - `apps/vela-gateway` now supports repeated placeholder input-audio append/commit cycles 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-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 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 that 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-ui` now exposes a cancel control for active push-to-talk-triggered mocked 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-gateway` now honors `response.cancel` during 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 - `apps/vela-protocol` now provides the shared WebSocket event contract for the UI and gateway
- backend framework choice is now concrete: Fastify - backend framework choice is now concrete: Fastify

1
docs/overview.md
View File

@@ -43,6 +43,7 @@ Vela is a fully local, voice-first assistant system with:
- browser-based PWA - browser-based PWA
- push-to-talk interaction - push-to-talk interaction
- current mocked vertical slice enters turns only through the placeholder push-to-talk shell
- transcript and response display - transcript and response display
- playback of streamed or returned audio - playback of streamed or returned audio

38
docs/protocol.md
View File

@@ -15,8 +15,8 @@ Current UI baseline:
- the browser opens a WebSocket directly to `/ws` - the browser opens a WebSocket directly to `/ws`
- the UI tracks connection status separately from gateway session status - the UI tracks connection status separately from gateway session status
- the UI can send `mocked.turn.trigger` after `session.ready` while connected to request one deterministic mocked turn for the active session
- the UI exposes a push-to-talk mic control shell that sends placeholder `input_audio.append` on press and `input_audio.commit` on release without capturing real audio - the UI exposes a push-to-talk mic control shell that sends placeholder `input_audio.append` on press and `input_audio.commit` on release without capturing real audio
- the push-to-talk shell is the only supported mocked turn entry path from the shipped UI
## WebSocket Message Envelope ## WebSocket Message Envelope
@@ -50,7 +50,7 @@ type ClientEvent =
#### Client event intent #### Client event intent
- `session.start` initializes a voice session without locking in transport or auth details yet - `session.start` initializes a voice session without locking in transport or auth details yet
- `mocked.turn.trigger` asks the gateway to run one obviously mocked, deterministic transcript/response turn - `mocked.turn.trigger` is a retired legacy event name that the gateway now rejects with a deterministic recoverable error
- `input_audio.append` carries a chunk of captured input audio as an encoded string - `input_audio.append` carries a chunk of captured input audio as an encoded string
- `input_audio.commit` marks the current buffered user turn as ready for downstream processing - `input_audio.commit` marks the current buffered user turn as ready for downstream processing
- `response.cancel` interrupts the active listen/think/speak flow - `response.cancel` interrupts the active listen/think/speak flow
@@ -59,15 +59,13 @@ type ClientEvent =
- on connect, the gateway creates an ephemeral in-memory session and emits `session.ready` plus `session.state` - on connect, the gateway creates an ephemeral in-memory session and emits `session.ready` plus `session.state`
- `session.start` is accepted as an idempotent session acknowledgment and re-sends readiness/state - `session.start` is accepted as an idempotent session acknowledgment and re-sends readiness/state
- `mocked.turn.trigger` is accepted only when no other mocked turn is already in flight for that session - `mocked.turn.trigger` is rejected deterministically with `error.code = unsupported_mocked_turn_trigger`
- a mocked turn emits deterministic `transcript.final`, `response.text.delta`, `response.completed`, and `session.state` events in protocol-valid order
- `input_audio.append` updates the ephemeral session record and moves the session to `listening` - `input_audio.append` updates the ephemeral session record and moves the session to `listening`
- each accepted `input_audio.append` emits one deterministic `transcript.partial` for the current placeholder turn - each accepted `input_audio.append` emits one deterministic `transcript.partial` for the current placeholder turn
- `input_audio.commit` emits exactly one deterministic `transcript.final` and then starts the same deterministic mocked assistant response stream used by `mocked.turn.trigger` - `input_audio.commit` emits exactly one deterministic `transcript.final` and then starts the deterministic mocked assistant response stream for that push-to-talk turn
- after a completed placeholder input cycle, the same socket can still send `mocked.turn.trigger` - after a completed placeholder input cycle, the same socket can start another placeholder push-to-talk turn without reconnecting
- `response.cancel` is safe to send even when no mocked turn is active - `response.cancel` is safe to send even when no mocked turn is active
- `response.cancel` stops any still-pending mocked turn events for the active turn and resets the minimal session state back to `idle` - `response.cancel` stops any still-pending mocked turn events for the active turn and resets the minimal session state back to `idle`
- a second mocked-turn trigger during an active mocked turn produces `error` with code `mocked_turn_in_flight`
- malformed JSON produces `error` with code `invalid_json` - malformed JSON produces `error` with code `invalid_json`
- invalid envelopes or unsupported client event names produce `error` with code `invalid_message` - invalid envelopes or unsupported client event names produce `error` with code `invalid_message`
- malformed WebSocket frames are rejected without crashing the gateway process - malformed WebSocket frames are rejected without crashing the gateway process
@@ -88,7 +86,6 @@ Notes:
- this UI state is transport-oriented and is separate from the shared gateway `session.state` payload - this UI state is transport-oriented and is separate from the shared gateway `session.state` payload
- `session.state` currently reflects the gateway session phase (`idle`, `listening`, `thinking`, `speaking`) - `session.state` currently reflects the gateway session phase (`idle`, `listening`, `thinking`, `speaking`)
- the UI disables the mocked-turn control until `session.ready` arrives, while disconnected, or while a mocked turn is already in flight
- the UI disables the mic control while disconnected, before `session.ready`, or while a mocked turn is already in flight - the UI disables the mic control while disconnected, before `session.ready`, or while a mocked turn is already in flight
- pressing the mic control sends one placeholder `input_audio.append` chunk and releasing it sends `input_audio.commit` - pressing the mic control sends one placeholder `input_audio.append` chunk and releasing it sends `input_audio.commit`
- while a placeholder push-to-talk turn is in progress, the UI renders the latest `transcript.partial` - while a placeholder push-to-talk turn is in progress, the UI renders the latest `transcript.partial`
@@ -126,26 +123,19 @@ type ServerEvent =
- `response.completed` marks the current assistant turn as done - `response.completed` marks the current assistant turn as done
- `error` is the minimal recoverable failure shape for both UI and gateway work - `error` is the minimal recoverable failure shape for both UI and gateway work
### Deterministic mocked turn sequence ### Legacy mocked turn trigger rejection
For this increment, `mocked.turn.trigger` produces one fixed interaction for the active session: For this increment, direct `mocked.turn.trigger` requests no longer start a mocked turn:
```text ```text
session.state(listening) mocked.turn.trigger
→ transcript.final("[mocked user] What is the current mocked vertical slice?") → error(code="unsupported_mocked_turn_trigger", message="mocked.turn.trigger is no longer supported; use input_audio.append and input_audio.commit instead.")
→ session.state(thinking)
→ session.state(speaking)
→ response.text.delta("[mocked assistant] ")
→ response.text.delta("This is a deterministic mocked response from the gateway vertical slice.")
→ response.completed
→ session.state(idle)
``` ```
Notes: Notes:
- the content is intentionally fixed and obviously mocked - this rejection is deterministic and recoverable
- no audio, STT, LLM, TTS, or external providers participate in this flow - the session remains available for the supported push-to-talk flow on the same socket
- `response.cancel` can stop the mocked turn early, suppress any later mocked response events for that turn, and return the session to `idle`
### Deterministic placeholder push-to-talk transcript and mocked response sequence ### Deterministic placeholder push-to-talk transcript and mocked response sequence
@@ -173,8 +163,8 @@ Safe deterministic edge cases for this mocked placeholder flow:
- commit without any prior append is accepted and emits `transcript.final("[mocked final] Placeholder push-to-talk transcript completed without appended audio.")` - commit without any prior append is accepted and emits `transcript.final("[mocked final] Placeholder push-to-talk transcript completed without appended audio.")`
- repeated appends during one placeholder turn are accepted and each append replaces the latest partial transcript with a chunk-count-based deterministic value - repeated appends during one placeholder turn are accepted and each append replaces the latest partial transcript with a chunk-count-based deterministic value
- after the final transcript, placeholder commit follows the same mocked `thinking → speaking → response.text.delta* → response.completed → idle` path as `mocked.turn.trigger` - after the final transcript, placeholder commit follows the deterministic mocked `thinking → speaking → response.text.delta* → response.completed → idle` path
- `response.cancel` can interrupt this mocked post-commit response path the same way it interrupts `mocked.turn.trigger`; already-rendered transcript or assistant text is not retracted - `response.cancel` can interrupt this mocked post-commit response path; already-rendered transcript or assistant text is not retracted
## Contract Scope for This Increment ## Contract Scope for This Increment
@@ -207,7 +197,7 @@ Current mocked-pipeline behavior:
- during an active mocked turn, `response.cancel` returns the session to `idle` immediately - during an active mocked turn, `response.cancel` returns the session to `idle` immediately
- any mocked turn timers that have not fired yet are dropped, so no later `response.text.delta` or `response.completed` events are emitted for the cancelled turn - any mocked turn timers that have not fired yet are dropped, so no later `response.text.delta` or `response.completed` events are emitted for the cancelled turn
- the same cancellation behavior applies when a mocked turn was started by `input_audio.commit` - the same cancellation behavior applies when a mocked turn was started by `input_audio.commit`
- once `idle` is restored, the same WebSocket session can start another mocked turn without reconnecting - once `idle` is restored, the same WebSocket session can start another placeholder push-to-talk turn without reconnecting
More general future-state expectations: More general future-state expectations: