@flowingspring/dsh-voco 0.3.6 → 0.3.8

package/lib/plugins/voice-local.js

@@ -5,6 +5,8 @@ import { existsSync } from "node:fs";
 import { dirname, join, resolve } from "node:path";
 import { loadEnvFile } from "node:process";
 import { fileURLToPath } from "node:url";
+import { credentialRef } from "@deepseek-ai/dsh-credentials";
+import { installSettingsSection, settingsNamespace } from "@deepseek-ai/dsh-settings";
 import { mkdtemp, readFile, rm } from "node:fs/promises";
 import { tmpdir } from "node:os";
 import { EdgeTTS } from "node-edge-tts";
@@ -70,6 +72,9 @@ var LocalSession = class {
 	async start() {
 		await this.backend.start((event) => this.receive(event));
 	}
+	beginManualUtterance() {
+		this.backend.beginManualUtterance();
+	}
 	appendAudio(audio) {
 		this.backend.appendAudio(audio);
 	}
@@ -312,6 +317,8 @@ var NodeSpeechBackend = class {
 	preRoll = [];
 	preRollBytes = 0;
 	active;
+	manual = false;
+	manualMaxed = false;
 	recognitionQueue = Promise.resolve();
 	synthesisQueue = Promise.resolve();
 	synthesisGeneration = 0;
@@ -343,8 +350,16 @@ var NodeSpeechBackend = class {
 		emit({ type: "ready" });
 		return Promise.resolve();
 	}
+	beginManualUtterance() {
+		if (this.closed) return;
+		this.active = void 0;
+		this.preRoll.splice(0);
+		this.preRollBytes = 0;
+		this.manual = true;
+		this.manualMaxed = false;
+	}
 	appendAudio(audio) {
-		if (this.closed || audio.byteLength === 0) return;
+		if (this.closed || audio.byteLength === 0 || this.manualMaxed) return;
 		if (audio.byteLength % 2 !== 0) {
 			this.emit?.({
 				type: "error",
@@ -367,14 +382,19 @@ var NodeSpeechBackend = class {
 			this.active.silenceBytes = 0;
 			if (!this.active.confirmed && this.active.voicedBytes >= this.bytesFor(this.config.minSpeechDurationMs ?? 250)) this.confirmUtterance();
 		} else this.active.silenceBytes += frame.byteLength;
-		if (!this.active.confirmed && this.active.silenceBytes >= this.bytesFor(this.config.minSpeechDurationMs ?? 250)) {
+		if (!this.manual && !this.active.confirmed && this.active.silenceBytes >= this.bytesFor(this.config.minSpeechDurationMs ?? 250)) {
 			this.active = void 0;
 			return;
 		}
-		if (this.active.silenceBytes >= this.bytesFor(this.config.silenceDurationMs) || this.active.totalBytes >= this.bytesFor(this.config.maxUtteranceMs)) this.finishUtterance();
+		if (!this.manual && this.active.silenceBytes >= this.bytesFor(this.config.silenceDurationMs) || this.active.totalBytes >= this.bytesFor(this.config.maxUtteranceMs)) {
+			this.finishUtterance();
+			if (this.manual) this.manualMaxed = true;
+		}
 	}
 	commitAudio() {
 		this.finishUtterance();
+		this.manual = false;
+		this.manualMaxed = false;
 	}
 	synthesize(responseId, text) {
 		const generation = this.synthesisGeneration;
@@ -433,6 +453,8 @@ var NodeSpeechBackend = class {
 		this.synthesisGeneration += 1;
 		this.synthesisResponses.clear();
 		this.active = void 0;
+		this.manual = false;
+		this.manualMaxed = false;
 		this.preRoll.splice(0);
 		await Promise.allSettled([this.recognitionQueue, this.synthesisQueue]);
 		this.emit?.({
@@ -539,8 +561,10 @@ function errorMessage(error) {
 /** Silence-gated SiliconFlow cloud ASR with Edge TTS. */
 const name = "voice-local";
 const inject = ["voice"];
+const VOICE_LOCAL_SETTINGS_NAMESPACE = settingsNamespace("voice-local");
+const SILICONFLOW_API_KEY_REF = credentialRef("SILICONFLOW_API_KEY");
 const Config = z.object({
-	apiKey: z.string(),
+	apiKey: z.string().role("secret"),
 	endpoint: z.string().default("https://api.siliconflow.cn/v1/audio/transcriptions"),
 	model: z.string().default("XingChenAGI/XingChenASR-V3.2-Ultra"),
 	interactionMode: z.union(["speech-shell", "frontend-agent"]).default("speech-shell"),
@@ -575,26 +599,34 @@ function loadProjectEnv() {
 	}
 }
 function apply(ctx, config = {}) {
-	loadProjectEnv();
+	if (ctx.get("credentials") === void 0) loadProjectEnv();
+	let current = () => config;
+	installSettingsSection(ctx, VOICE_LOCAL_SETTINGS_NAMESPACE, Config, config, {
+		setSource: (source) => {
+			current = source;
+		},
+		onChange: () => {}
+	});
 	return ctx.voice.registerProvider({
 		id: "local",
 		available: () => true,
 		connect: async ({ voiceSessionId, emit }) => {
+			const active = current();
 			const session = new LocalSession(new NodeSpeechBackend({
-				apiKey: config.apiKey ?? process.env.SILICONFLOW_API_KEY ?? "",
-				endpoint: config.endpoint ?? "https://api.siliconflow.cn/v1/audio/transcriptions",
-				model: config.model ?? "XingChenAGI/XingChenASR-V3.2-Ultra",
-				requestTimeoutMs: config.requestTimeoutMs ?? 6e4,
-				inputSampleRate: config.inputSampleRate ?? 16e3,
-				outputSampleRate: config.outputSampleRate ?? 48e3,
-				ttsRate: config.ttsRate ?? "+20%",
-				silenceDurationMs: config.silenceDurationMs ?? 1500,
-				speechThreshold: config.speechThreshold ?? .015,
-				minSpeechDurationMs: config.minSpeechDurationMs ?? 250,
-				preRollMs: config.preRollMs ?? 400,
-				trailingSilenceMs: config.trailingSilenceMs ?? 200,
-				maxUtteranceMs: config.maxUtteranceMs ?? 6e4
-			}), emit, voiceSessionId, config.interactionMode ?? "speech-shell");
+				apiKey: await resolveSiliconFlowApiKey(ctx, active),
+				endpoint: active.endpoint ?? "https://api.siliconflow.cn/v1/audio/transcriptions",
+				model: active.model ?? "XingChenAGI/XingChenASR-V3.2-Ultra",
+				requestTimeoutMs: active.requestTimeoutMs ?? 6e4,
+				inputSampleRate: active.inputSampleRate ?? 16e3,
+				outputSampleRate: active.outputSampleRate ?? 48e3,
+				ttsRate: active.ttsRate ?? "+20%",
+				silenceDurationMs: active.silenceDurationMs ?? 1500,
+				speechThreshold: active.speechThreshold ?? .015,
+				minSpeechDurationMs: active.minSpeechDurationMs ?? 250,
+				preRollMs: active.preRollMs ?? 400,
+				trailingSilenceMs: active.trailingSilenceMs ?? 200,
+				maxUtteranceMs: active.maxUtteranceMs ?? 6e4
+			}), emit, voiceSessionId, active.interactionMode ?? "speech-shell");
 			try {
 				await session.start();
 				return session;
@@ -605,5 +637,12 @@ function apply(ctx, config = {}) {
 		}
 	});
 }
+/** Resolve the key per connection so a settings-page write reaches the next attempt. */
+async function resolveSiliconFlowApiKey(ctx, config) {
+	if (config.apiKey !== void 0 && config.apiKey.trim() !== "") return config.apiKey;
+	const credentials = ctx.get("credentials");
+	if (credentials !== void 0) return (await credentials.resolve(SILICONFLOW_API_KEY_REF))?.value ?? "";
+	return process.env.SILICONFLOW_API_KEY ?? "";
+}
 //#endregion
-export { Config, LocalSession, NodeSpeechBackend, SiliconFlowAsr, apply, inject, name, pcm16MonoWav };
+export { Config, LocalSession, NodeSpeechBackend, SILICONFLOW_API_KEY_REF, SiliconFlowAsr, VOICE_LOCAL_SETTINGS_NAMESPACE, apply, inject, name, pcm16MonoWav, resolveSiliconFlowApiKey };

package/lib/plugins/voice-web.js

@@ -213,7 +213,8 @@ async function attach(ctx, socket, rawUrl) {
 				return;
 			}
 			const type = control.type;
-			if (type === "audio.commit") voice.commitAudio(session.id);
+			if (type === "audio.push-to-talk.start") voice.beginManualUtterance(session.id);
+			else if (type === "audio.commit") voice.commitAudio(session.id);
 			else if (type === "response.interrupt") voice.interruptResponse(session.id);
 			else if (type === "playback.ended") voice.playbackEnded(session.id);
 			else if (type === "task.cancel") {

package/lib/plugins/voice.js

@@ -169,6 +169,10 @@ var VoiceRuntime = class extends Service {
 	appendAudio(id, audio) {
 		this.requireAttachedSession(id).provider.appendAudio(audio);
 	}
+	/** Begin a manually committed push-to-talk utterance. */
+	beginManualUtterance(id) {
+		this.requireAttachedSession(id).provider.beginManualUtterance();
+	}
 	/**
 	* Commit current microphone input.
 	* @param id - voice session.

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@flowingspring/dsh-voco",
   "description": "Persistent voice conversations for DSH with cloud speech recognition, Edge TTS, and background Agent delegation",
-  "version": "0.3.6",
+  "version": "0.3.8",
   "keywords": [
     "deepseek-harness",
     "dsh",
@@ -65,6 +65,8 @@
         "@deepseek-ai/dsh-client-connection",
         "@deepseek-ai/dsh-client-locale",
         "@deepseek-ai/dsh-client-runtime",
+        "@deepseek-ai/dsh-client-ui-settings",
+        "@deepseek-ai/dsh-client-ui-settings-plugins",
         "@deepseek-ai/dsh-client-ui-conversation",
         "@deepseek-ai/dsh-client-ui-layout",
         "@deepseek-ai/dsh-client-ui-sidebar"
@@ -86,15 +88,19 @@
     "@deepseek-ai/dsh-client-connection": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-locale": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-runtime": "^0.1.1-rc.1",
+    "@deepseek-ai/dsh-client-ui-settings": "^0.1.1-rc.2",
+    "@deepseek-ai/dsh-client-ui-settings-plugins": "^0.1.1-rc.2",
     "@deepseek-ai/dsh-client-ui-conversation": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-ui-layout": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-ui-sidebar": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-ui-slots": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-host-webserver": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-invariants": "^0.1.1-rc.1",
+    "@deepseek-ai/dsh-credentials": "^0.1.1-rc.2",
     "@deepseek-ai/dsh-llm": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-session": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-session-persistence": "^0.1.1-rc.1",
+    "@deepseek-ai/dsh-settings": "^0.1.1-rc.2",
     "@deepseek-ai/dsh-system-prompt": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-tools": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-workspace": "^0.1.1-rc.1",
@@ -109,15 +115,19 @@
     "@deepseek-ai/dsh-client-connection": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-locale": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-runtime": "^0.1.1-rc.1",
+    "@deepseek-ai/dsh-client-ui-settings": "^0.1.1-rc.2",
+    "@deepseek-ai/dsh-client-ui-settings-plugins": "^0.1.1-rc.2",
     "@deepseek-ai/dsh-client-ui-conversation": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-ui-layout": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-ui-sidebar": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-client-ui-slots": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-host-webserver": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-invariants": "^0.1.1-rc.1",
+    "@deepseek-ai/dsh-credentials": "^0.1.1-rc.2",
     "@deepseek-ai/dsh-llm": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-session": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-session-persistence": "^0.1.1-rc.1",
+    "@deepseek-ai/dsh-settings": "^0.1.1-rc.2",
     "@deepseek-ai/dsh-system-prompt": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-tools": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-workspace": "^0.1.1-rc.1",
@@ -125,11 +135,11 @@
     "@types/ws": "^8.18.1",
     "react": "^18.2.0",
     "react-dom": "^18.2.0",
+    "@flowingspring/dsh-voice": "0.1.2",
     "@flowingspring/dsh-voice-local": "0.1.0",
-    "@flowingspring/dsh-client-ui-voice": "0.1.2",
     "@flowingspring/dsh-llm-tool-call-compat": "0.1.0",
-    "@flowingspring/dsh-voice": "0.1.2",
     "@flowingspring/dsh-voice-web": "0.1.2",
+    "@flowingspring/dsh-client-ui-voice": "0.1.2",
     "@flowingspring/dsh-voice-assistant": "0.1.2"
   }
 }