@doufunao123/asset-gateway 0.16.0 → 0.17.0

package/dist/index.js

@@ -1,7 +1,7 @@
 #!/usr/bin/env node
 
 // src/index.ts
-import { Command as Command10 } from "commander";
+import { Command as Command9 } from "commander";
 
 // src/commands/auth.ts
 import { existsSync as existsSync2, unlinkSync } from "fs";
@@ -415,12 +415,11 @@ var SCHEMAS = {
           "--output-dir": { type: "string", default: "." }
         }
       },
-      audio: {
-        description: "Generate audio from a text prompt",
+      sfx: {
+        description: "Generate sound effects (impacts, footsteps, UI sounds, ambience)",
         params: {
           "--prompt": { type: "string", required: true },
-          "--type": { type: "string", description: "bgm | sfx" },
-          "--duration": { type: "number", description: "Seconds" },
+          "--duration": { type: "number", required: true, description: "Duration in seconds (1-5s for short SFX, 5-15s for ambience, max 30s)" },
           "--output-dir": { type: "string", default: "." }
         }
       },
@@ -435,10 +434,11 @@ var SCHEMAS = {
         }
       },
       tts: {
-        description: "Text-to-speech via MOSS-TTS-Nano (self-hosted)",
+        description: "Text-to-speech via Gemini 3.1 Flash TTS",
         params: {
           "--prompt": { type: "string", required: true },
-          "--input": { type: "string", description: "Reference audio file for voice cloning (local path or URL)" },
+          "--voice": { type: "string", description: "Prebuilt voice name (default: Kore)" },
+          "--speakers": { type: "string", description: `Multi-speaker config JSON, e.g. '{"Name1":"Puck","Name2":"Kore"}'` },
           "--output-dir": { type: "string", default: "." }
         }
       },
@@ -650,22 +650,6 @@ var SCHEMAS = {
       }
     }
   },
-  voice: {
-    description: "Qwen3-TTS voice designs (design / list / delete)",
-    subcommands: {
-      design: {
-        description: "Design voice from text",
-        params: {
-          "--prompt": { type: "string", required: true },
-          "--preview-text": { type: "string", required: true },
-          "--name": { type: "string", required: true },
-          "--target-model": { type: "string" }
-        }
-      },
-      list: { description: "List designed voices", params: { "--type": { type: "string" } } },
-      delete: { description: "Delete designed voice", params: { "<voice-id>": { type: "string", required: true }, "--type": { type: "string" } } }
-    }
-  },
   upload: {
     description: "Upload and list gateway assets",
     subcommands: {
@@ -739,7 +723,7 @@ import { existsSync as existsSync3, mkdirSync as mkdirSync2, readFileSync as rea
 import { dirname as dirname2, extname, join as join2 } from "path";
 import { Command as Command3 } from "commander";
 function inferExtension(assetType) {
-  const map = { image: "png", audio: "mp3", music: "mp3", tts: "mp3", video: "mp4", model3d: "glb", text: "txt", sprite: "png", world: "spz" };
+  const map = { image: "png", audio: "mp3", sfx: "mp3", music: "mp3", tts: "mp3", video: "mp4", model3d: "glb", text: "txt", sprite: "png", world: "spz" };
   return map[assetType] ?? "bin";
 }
 function inferExtFromResult(result) {
@@ -909,23 +893,23 @@ function createGenerateCommand() {
     })
   );
   command.addCommand(
-    new Command3("audio").description("Generate audio from a text prompt").requiredOption("--prompt <text>", "Audio description prompt").option("--type <type>", "Audio type: bgm or sfx").option("--duration <seconds>", "Duration in seconds").option("--output-dir <dir>", "Directory to save output", ".").action(async function(options) {
+    new Command3("sfx").description("Generate sound effects (short audio clips: impacts, footsteps, UI sounds, ambience)").requiredOption("--prompt <text>", "Sound effect description").requiredOption("--duration <seconds>", "Duration in seconds (1-5s for short SFX, 5-15s for ambience, max 30s)").option("--output-dir <dir>", "Directory to save output", ".").action(async function(options) {
       try {
         const ctx = createContext(this);
+        const params = {
+          duration_seconds: Number(options.duration)
+        };
         const body = {
           asset_type: "audio",
-          prompt: options.prompt
+          prompt: options.prompt,
+          params
         };
-        const params = {};
-        if (options.type) params.audio_type = options.type;
-        if (options.duration) params.duration_seconds = Number(options.duration);
-        if (Object.keys(params).length > 0) body.params = params;
         const data = await ctx.client.post("/api/generate", body);
         const localPath = await saveOutput(data, "audio", options.outputDir);
         if (localPath) data.local_path = localPath;
-        printSuccess("generate.audio", data, ctx);
+        printSuccess("generate.sfx", data, ctx);
       } catch (error2) {
-        printError("generate.audio", error2);
+        printError("generate.sfx", error2);
       }
     })
   );
@@ -955,23 +939,17 @@ function createGenerateCommand() {
     })
   );
   command.addCommand(
-    new Command3("tts").description("Text-to-speech via MOSS-TTS-Nano (self-hosted)").requiredOption("--prompt <text>", "Text to synthesize").option("--input <path>", "Reference audio file for voice cloning (local path or URL)").option("--output-dir <dir>", "Directory to save output", ".").action(async function(options) {
+    new Command3("tts").description("Text-to-speech via Gemini 3.1 Flash TTS").requiredOption("--prompt <text>", "Text to synthesize").option("--voice <name>", "Prebuilt voice name (default: Kore)").option("--speakers <json>", `Multi-speaker config JSON, e.g. '{"Name1":"Puck","Name2":"Kore"}'`).option("--output-dir <dir>", "Directory to save output", ".").action(async function(options) {
       try {
         const ctx = createContext(this);
         const body = {
           asset_type: "tts",
           prompt: options.prompt
         };
-        if (options.input) {
-          if (existsSync3(options.input)) {
-            const ext = extname(options.input).toLowerCase();
-            const mime = ext === ".wav" ? "audio/wav" : ext === ".mp3" ? "audio/mpeg" : "audio/wav";
-            const b64 = readFileSync2(options.input).toString("base64");
-            body.input_file = `data:${mime};base64,${b64}`;
-          } else {
-            body.input_file = options.input;
-          }
-        }
+        const params = {};
+        if (options.voice) params.voice = options.voice;
+        if (options.speakers) params.speakers = JSON.parse(options.speakers);
+        if (Object.keys(params).length > 0) body.params = params;
         const data = await ctx.client.post("/api/generate", body);
         const localPath = await saveOutput(data, "tts", options.outputDir);
         if (localPath) data.local_path = localPath;
@@ -1648,106 +1626,8 @@ function createUploadCommand() {
   return command;
 }
 
-// src/commands/voice.ts
-import { Command as Command9 } from "commander";
-function withVoiceType(path, type) {
-  if (!type) {
-    return path;
-  }
-  const params = new URLSearchParams({ type });
-  return `${path}?${params.toString()}`;
-}
-function createVoiceCommand() {
-  const command = new Command9("voice").description("Manage Qwen3-TTS voice designs");
-  command.addCommand(
-    new Command9("design").description("Create a synthetic voice and save the preview audio file").requiredOption("--prompt <text>", "Voice description prompt").requiredOption("--preview-text <text>", "Preview text for the generated sample").requiredOption("--name <name>", "Name for the designed voice").option("--target-model <model>", "Voice design model, e.g. qwen3-tts-vd-2026-01-26").option("--output-dir <dir>", "Directory to save preview audio", ".").action(async function(options) {
-      try {
-        const ctx = createContext(this);
-        const body = {
-          voice_prompt: options.prompt,
-          preview_text: options.previewText,
-          name: options.name
-        };
-        if (options.targetModel) body.target_model = options.targetModel;
-        const data = await ctx.client.post("/api/voice/design", body);
-        const designData = data?.data;
-        const b64Audio = designData?.preview_audio_data;
-        if (b64Audio) {
-          const { mkdirSync: mkdirSync5, writeFileSync: writeFileSync5 } = await import("fs");
-          const { join: join5 } = await import("path");
-          mkdirSync5(options.outputDir, { recursive: true });
-          const buf = Buffer.from(b64Audio, "base64");
-          const outPath = join5(options.outputDir, `${options.name}_preview.wav`);
-          writeFileSync5(outPath, buf);
-          data.preview_audio_path = outPath;
-        }
-        printSuccess("voice.design", data, ctx);
-      } catch (error2) {
-        printError("voice.design", error2);
-      }
-    })
-  );
-  command.addCommand(
-    new Command9("synthesize").description("Synthesize speech using a designed voice").requiredOption("--voice <name>", "Voice name from voice design").requiredOption("--text <text>", "Text to synthesize").option("--model <model>", "TTS model (default: qwen3-tts-vd-realtime-2025-12-16)").option("--language <lang>", "Language: zh, en, ja, etc. (default: Auto)").option("--output-dir <dir>", "Directory to save output", ".").action(async function(options) {
-      try {
-        const ctx = createContext(this);
-        const body = {
-          voice: options.voice,
-          text: options.text
-        };
-        if (options.model) body.model = options.model;
-        if (options.language) body.language = options.language;
-        const data = await ctx.client.post("/api/voice/synthesize", body);
-        const result = data.data;
-        const output2 = result?.output;
-        const audio = output2?.audio;
-        const audioUrl = audio?.url ?? output2?.url;
-        if (audioUrl && options.outputDir) {
-          const { mkdirSync: mkdirSync5, writeFileSync: writeFileSync5 } = await import("fs");
-          const { join: join5 } = await import("path");
-          mkdirSync5(options.outputDir, { recursive: true });
-          const resp = await fetch(audioUrl);
-          if (resp.ok) {
-            const buf = Buffer.from(await resp.arrayBuffer());
-            const outPath = join5(options.outputDir, `tts_${Date.now()}.wav`);
-            writeFileSync5(outPath, buf);
-            data.local_path = outPath;
-          }
-        }
-        printSuccess("voice.synthesize", data, ctx);
-      } catch (error2) {
-        printError("voice.synthesize", error2);
-      }
-    })
-  );
-  command.addCommand(
-    new Command9("list").description("List custom designed voices").option("--type <type>", "Voice type: vd").action(async function(options) {
-      try {
-        const ctx = createContext(this);
-        const data = await ctx.client.get(withVoiceType("/api/voice/list", options.type));
-        printSuccess("voice.list", data, ctx);
-      } catch (error2) {
-        printError("voice.list", error2);
-      }
-    })
-  );
-  command.addCommand(
-    new Command9("delete").description("Delete a custom designed voice").argument("<voice-id>", "Voice ID to delete").option("--type <type>", "Voice type: vd").action(async function(voiceId, options) {
-      try {
-        const ctx = createContext(this);
-        const path = withVoiceType(`/api/voice/${encodeURIComponent(voiceId)}`, options.type);
-        const data = await ctx.client.delete(path);
-        printSuccess("voice.delete", data, ctx);
-      } catch (error2) {
-        printError("voice.delete", error2);
-      }
-    })
-  );
-  return command;
-}
-
 // src/index.ts
-var program = new Command10().name("asset-gateway").description("Universal asset generation gateway CLI").version(CLI_VERSION).option(
+var program = new Command9().name("asset-gateway").description("Universal asset generation gateway CLI").version(CLI_VERSION).option(
   "--gateway-url <url>",
   `Gateway URL (default: $ASSET_GATEWAY_URL, auth config, or ${DEFAULT_GATEWAY_URL})`
 ).option("--token <token>", "API token for authentication").option("--human", "Human-readable output instead of JSON").option("--fields <fields>", "Comma-separated list of output fields");
@@ -1757,7 +1637,6 @@ program.addCommand(createProcessCommand());
 program.addCommand(createProcess3dCommand());
 program.addCommand(createProviderCommand());
 program.addCommand(createUploadCommand());
-program.addCommand(createVoiceCommand());
 program.addCommand(createJobCommand());
 program.addCommand(createDescribeCommand());
 await program.parseAsync(process.argv);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@doufunao123/asset-gateway",
-  "version": "0.16.0",
+  "version": "0.17.0",
   "description": "Universal asset generation gateway CLI",
   "type": "module",
   "bin": {