mikkel/nexus
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

feat(36-03): wire voiceMode through chat stream, mount voice routes, remove old transcribe

Browse source 
- server/src/routes/chat.ts: destructure voiceMode from req.body in stream endpoint
- server/src/routes/chat.ts: inject dual-output system prompt when voiceMode=full_voice (VPIPE-06)
- server/src/routes/chat.ts: persist voiceMode to messageType column (voice_full/voice_input)
- server/src/routes/chat-files.ts: remove old inline /transcribe endpoint (lines 297-386)
- server/src/app.ts: import and mount voiceRoutes() after nexusSettingsRoutes()
This commit is contained in:
Nexus Dev 2026-04-04 01:39:31 +00:00
parent bf757509ef
commit f23c4218aa
3 changed files with 22 additions and 92 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
server/src/app.ts
View file

@ -32,6 +32,7 @@ import { assistantMemoryRoutes } from "./routes/assistant-memory.js";
import { assistantHandoffRoutes } from "./routes/assistant-handoff.js";
import { chatFileRoutes } from "./routes/chat-files.js";
import { nexusSettingsRoutes } from "./routes/nexus-settings.js";
import { voiceRoutes } from "./routes/voice.js";
import { pluginRoutes } from "./routes/plugins.js";
import { pluginUiStaticRoutes } from "./routes/plugin-ui-static.js";
import { applyUiBranding } from "./ui-branding.js";
@ -174,6 +175,7 @@ export async function createApp(
api.use(assistantHandoffRoutes(db));
api.use(chatFileRoutes(db, opts.storageService));
api.use(nexusSettingsRoutes());
api.use(voiceRoutes());
const hostServicesDisposers = new Map<string, () => void>();
const workerManager = createPluginWorkerManager();
const pluginRegistry = pluginRegistryService(db);

91
server/src/routes/chat-files.ts
View file

@ -294,96 +294,5 @@ export function chatFileRoutes(db: Db, storage: StorageService) {
res.json(updated);
});
// POST /transcribe — Transcribe audio via local Whisper
const audioUpload = multer({
storage: multer.memoryStorage(),
limits: { fileSize: MAX_ATTACHMENT_BYTES, files: 1 },
});
async function runAudioUpload(
upload: ReturnType<typeof multer>,
req: Request,
res: Response,
) {
await new Promise<void>((resolve, reject) => {
upload.single("audio")(req, res, (err: unknown) => {
if (err) reject(err);
else resolve();
});
});
}
router.post("/transcribe", async (req, res) => {
assertBoard(req);
try {
await runAudioUpload(audioUpload, req, res);
} catch (err) {
if (err instanceof multer.MulterError) {
res.status(400).json({ error: err.message });
return;
}
throw err;
}
const file = (req as Request & { file?: { buffer: Buffer; mimetype: string } }).file;
if (!file) {
res.status(400).json({ error: "Missing audio field" });
return;
}
// Write to temp file
const { writeFile, unlink } = await import("node:fs/promises");
const { tmpdir } = await import("node:os");
const tmpPath = path.join(tmpdir(), `nexus-audio-${Date.now()}.webm`);
try {
await writeFile(tmpPath, file.buffer);
// Try whisper CLI (whisper.cpp or openai-whisper)
const { promisify } = await import("node:util");
const { execFile: execFileCb } = await import("node:child_process");
const execFileAsync = promisify(execFileCb);
try {
// Try whisper.cpp first (outputs transcription to stdout with --no-timestamps)
const { stdout } = await execFileAsync("whisper-cpp", [
"--model", "base.en",
"--file", tmpPath,
"--no-timestamps",
"--output-txt",
], { timeout: 30000 });
res.json({ text: stdout.trim() });
} catch {
try {
// Fallback: openai-whisper Python CLI
const { stdout } = await execFileAsync("whisper", [
tmpPath,
"--model", "base.en",
"--output_format", "txt",
"--output_dir", tmpdir(),
], { timeout: 60000 });
// whisper CLI writes to a .txt file alongside the input
const txtPath = tmpPath.replace(/\.webm$/, ".txt");
try {
const { readFile } = await import("node:fs/promises");
const text = await readFile(txtPath, "utf-8");
await unlink(txtPath).catch(() => {});
res.json({ text: text.trim() });
} catch {
// Parse stdout as fallback
res.json({ text: stdout.trim() });
}
} catch {
res.status(503).json({
error: "Whisper not available. Install whisper-cpp or openai-whisper for voice input.",
});
}
}
} finally {
await unlink(tmpPath).catch(() => {});
}
});
return router;
}

21
server/src/routes/chat.ts
View file

@ -90,7 +90,9 @@ export function chatRoutes(db: Db): Router {
// POST /api/conversations/:id/stream -- SSE streaming endpoint (CHAT-01, PERF-02)
router.post("/conversations/:id/stream", async (req, res) => {
assertBoard(req);
const { content, agentId } = req.body;
const { content, agentId, voiceMode } = req.body as {
content: string; agentId?: string; voiceMode?: "text" | "voice_input" | "full_voice";
};
if (!content || typeof content !== "string") {
res.status(400).json({ error: "content is required" });
return;
@ -139,6 +141,20 @@ export function chatRoutes(db: Db): Router {
// Add the new user message
messagesWithMemory.push({ role: "user", content });
// Inject dual-output formatting prompt when voice mode is full_voice (VPIPE-06)
if (voiceMode === "full_voice") {
messagesWithMemory.push({
role: "system",
content: [
"Format your response with EXACTLY these two labeled sections:",
"",
"SPOKEN: [Natural speech prose only. No markdown. No bullet points. No code blocks. Max 2-3 sentences for spoken delivery.]",
"",
"DETAILED: [Your full response with all detail, code blocks, and markdown formatting.]",
].join("\n"),
});
}
// Set SSE headers and flush BEFORE any generation (PERF-02)
res.setHeader("Content-Type", "text/event-stream");
res.setHeader("Cache-Control", "no-cache");
@ -168,6 +184,9 @@ export function chatRoutes(db: Db): Router {
role: "assistant",
content: fullContent.trim(),
agentId: agentId || undefined,
messageType: voiceMode === "full_voice" ? "voice_full"
: voiceMode === "voice_input" ? "voice_input"
: undefined,
});
res.write(`data: ${JSON.stringify({ type: "done", messageId: message.id, content: fullContent.trim() })}\n\n`);