@doufunao123/asset-gateway 0.14.3 → 0.16.0

package/dist/index.js

@@ -435,15 +435,10 @@ var SCHEMAS = {
         }
       },
       tts: {
-        description: "TTS: default Qwen3-TTS (voice/language/instructions). ElevenLabs: --provider elevenlabs --voice-id <id>",
+        description: "Text-to-speech via MOSS-TTS-Nano (self-hosted)",
         params: {
           "--prompt": { type: "string", required: true },
-          "--voice": { type: "string", description: "Qwen voice name or custom id", default: "Cherry" },
-          "--voice-id": { type: "string", description: "ElevenLabs voice id (with --provider elevenlabs)" },
-          "--language": { type: "string", default: "Auto" },
-          "--model": { type: "string", default: "qwen3-tts-flash" },
-          "--instructions": { type: "string", description: "Instruct-model style control" },
-          "--provider": { type: "string", description: "qwen_tts | elevenlabs" },
+          "--input": { type: "string", description: "Reference audio file for voice cloning (local path or URL)" },
           "--output-dir": { type: "string", default: "." }
         }
       },
@@ -472,19 +467,15 @@ var SCHEMAS = {
         }
       },
       sprite: {
-        description: "Generate character animation (Veo AI video + frame extraction)",
+        description: "Generate character animation spritesheet (AutoSprite)",
         params: {
           "--prompt": { type: "string", required: true, description: "Character description" },
           "--input": { type: "string", description: "Reference image path or URL" },
-          "--animation-type": { type: "string", default: "walk" },
-          "--direction": { type: "string", default: "right" },
-          "--view": { type: "string", default: "auto", description: "Camera view: auto, side, front, back, three-quarter, none" },
-          "--framing": { type: "string", default: "full-body", description: "Framing: full-body, waist-up, close-up, none" },
-          "--background": { type: "string", default: "auto", description: "Background: auto, white, none, or free text" },
-          "--duration": { type: "number", default: 2, description: "Video duration in seconds" },
-          "--style": { type: "string" },
-          "--output-format": { type: "string", default: "spritesheet", description: "spritesheet | gif" },
-          "--fps": { type: "number", default: 8, description: "GIF frame rate" },
+          "--animation-type": { type: "string", default: "walk", description: "walk, run, idle, jump, attack, death, cast, dance, wave, interact, or custom text" },
+          "--style": { type: "string", description: "Art style: 16-bit, hd-pixel, isometric, retro-8bit, anime, chibi, painterly, vector" },
+          "--frame-count": { type: "number", default: 8, description: "Number of animation frames" },
+          "--frame-size": { type: "number", default: 256, description: "Frame size in pixels (square)" },
+          "--is-humanoid": { type: "boolean", default: true },
           "--output-dir": { type: "string", default: "." }
         }
       },
@@ -660,16 +651,8 @@ var SCHEMAS = {
     }
   },
   voice: {
-    description: "Qwen3-TTS custom voices (clone / design / list / delete)",
+    description: "Qwen3-TTS voice designs (design / list / delete)",
     subcommands: {
-      clone: {
-        description: "Clone from audio sample",
-        params: {
-          "--audio": { type: "string", required: true },
-          "--name": { type: "string", required: true },
-          "--target-model": { type: "string" }
-        }
-      },
       design: {
         description: "Design voice from text",
         params: {
@@ -679,8 +662,8 @@ var SCHEMAS = {
           "--target-model": { type: "string" }
         }
       },
-      list: { description: "List custom voices", params: { "--type": { type: "string" } } },
-      delete: { description: "Delete by voice id", params: { "<voice-id>": { type: "string", required: true }, "--type": { 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: {
@@ -972,28 +955,23 @@ function createGenerateCommand() {
     })
   );
   command.addCommand(
-    new Command3("tts").description(
-      "Text-to-speech: default Qwen3-TTS; use --provider elevenlabs --voice-id for ElevenLabs"
-    ).requiredOption("--prompt <text>", "Text to synthesize").option("--voice <name>", "Qwen voice name or custom voice id", "Cherry").option(
-      "--voice-id <id>",
-      "ElevenLabs voice_id (use with --provider elevenlabs; routes to TTS API)"
-    ).option("--language <lang>", "Language hint: Auto, Chinese, English, Japanese, etc.", "Auto").option("--model <model>", "Model id (default: auto-detect from voice; qwen3-tts-flash for built-in voices)").option("--instructions <text>", "Natural language speaking instructions (for instruct models)").option("--provider <id>", "qwen_tts | elevenlabs | voicebox").option("--profile-id <id>", "VoiceBox profile_id (use with --provider voicebox)").option("--output-dir <dir>", "Directory to save output", ".").action(async function(options) {
+    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) {
       try {
         const ctx = createContext(this);
-        const params = {
-          voice: options.voice,
-          language_type: options.language
-        };
-        if (options.instructions) params.instructions = options.instructions;
-        if (options.voiceId) params.voice_id = options.voiceId;
-        if (options.profileId) params.profile_id = options.profileId;
         const body = {
           asset_type: "tts",
-          prompt: options.prompt,
-          params
+          prompt: options.prompt
         };
-        if (options.model) body.model = options.model;
-        if (options.provider) body.provider = options.provider;
+        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 data = await ctx.client.post("/api/generate", body);
         const localPath = await saveOutput(data, "tts", options.outputDir);
         if (localPath) data.local_path = localPath;
@@ -1052,19 +1030,15 @@ function createGenerateCommand() {
     })
   );
   command.addCommand(
-    new Command3("sprite").description("Generate character animation (Veo AI video + frame extraction)").requiredOption("--prompt <text>", "Character description").option("--input <path>", "Reference image for character consistency (local path or URL)").option("--animation-type <type>", "Animation type (idle, walk, run, attack, death, jump, cast, dance, or any custom)", "walk").option("--direction <dir>", "Facing direction: front, left, right, back", "front").option("--view <view>", "Camera view angle: auto, side, front, back, three-quarter, none", "auto").option("--framing <framing>", "Framing: full-body, waist-up, close-up, none", "full-body").option("--background <bg>", "Background: auto, white, none, or free text (e.g. 'forest clearing')", "auto").option("--duration <n>", "Video duration in seconds (1-15)", "2").option("--style <style>", "Visual style (e.g. pixel art, hand-drawn, chibi)").option("--output-format <fmt>", "Output format: spritesheet or gif", "spritesheet").option("--fps <n>", "GIF frame rate", "8").option("--output-dir <dir>", "Directory to save output", ".").action(async function(options) {
+    new Command3("sprite").description("Generate character animation spritesheet (AutoSprite)").requiredOption("--prompt <text>", "Character description").option("--input <path>", "Reference image for character consistency (local path or URL)").option("--animation-type <type>", "Animation type: walk, run, idle, jump, attack, death, cast, dance, wave, interact, or custom text", "walk").option("--style <style>", "Art style: 16-bit, hd-pixel, isometric, retro-8bit, anime, chibi, painterly, vector, or any text").option("--frame-count <n>", "Number of animation frames", "8").option("--frame-size <n>", "Frame size in pixels (square)", "256").option("--is-humanoid", "Character is humanoid (default true)", true).option("--output-dir <dir>", "Directory to save output", ".").action(async function(options) {
       try {
         const ctx = createContext(this);
         const params = {
           animation_type: options.animationType,
-          direction: options.direction,
-          duration: Number(options.duration),
-          output_format: options.outputFormat,
-          fps: Number(options.fps)
+          frame_count: Number(options.frameCount),
+          frame_size: Number(options.frameSize),
+          is_humanoid: options.isHumanoid
         };
-        if (options.view && options.view !== "auto") params.view = options.view;
-        if (options.framing && options.framing !== "full-body") params.framing = options.framing;
-        if (options.background && options.background !== "auto") params.background = options.background;
         if (options.style) params.style = options.style;
         let inputFile = options.input;
         if (inputFile && existsSync3(inputFile)) {
@@ -1675,33 +1649,7 @@ function createUploadCommand() {
 }
 
 // src/commands/voice.ts
-import { existsSync as existsSync5, readFileSync as readFileSync4 } from "fs";
-import { extname as extname2 } from "path";
 import { Command as Command9 } from "commander";
-function inferAudioMime(filePath) {
-  const extension = extname2(filePath).toLowerCase();
-  const map = {
-    ".mp3": "audio/mpeg",
-    ".wav": "audio/wav",
-    ".pcm": "audio/pcm",
-    ".opus": "audio/opus",
-    ".ogg": "audio/ogg",
-    ".m4a": "audio/mp4",
-    ".aac": "audio/aac",
-    ".flac": "audio/flac"
-  };
-  return map[extension] ?? "application/octet-stream";
-}
-function readAudioAsBase64(filePath) {
-  if (!existsSync5(filePath)) {
-    throw configError(`Audio file not found: ${filePath}`);
-  }
-  const bytes = readFileSync4(filePath);
-  return {
-    audio_base64: bytes.toString("base64"),
-    audio_mime: inferAudioMime(filePath)
-  };
-}
 function withVoiceType(path, type) {
   if (!type) {
     return path;
@@ -1710,26 +1658,9 @@ function withVoiceType(path, type) {
   return `${path}?${params.toString()}`;
 }
 function createVoiceCommand() {
-  const command = new Command9("voice").description("Manage Qwen3-TTS custom voices");
+  const command = new Command9("voice").description("Manage Qwen3-TTS voice designs");
   command.addCommand(
-    new Command9("clone").description("Clone a voice from an audio sample").requiredOption("--audio <path>", "Reference audio file path").requiredOption("--name <name>", "Name for the cloned voice").option("--target-model <model>", "Voice cloning model, e.g. qwen3-tts-vc-2026-01-22").action(async function(options) {
-      try {
-        const ctx = createContext(this);
-        const audio = readAudioAsBase64(options.audio);
-        const body = {
-          ...audio,
-          name: options.name
-        };
-        if (options.targetModel) body.target_model = options.targetModel;
-        const data = await ctx.client.post("/api/voice/clone", body);
-        printSuccess("voice.clone", data, ctx);
-      } catch (error2) {
-        printError("voice.clone", error2);
-      }
-    })
-  );
-  command.addCommand(
-    new Command9("design").description("Create a synthetic voice from a text description").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").action(async function(options) {
+    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 = {
@@ -1739,6 +1670,17 @@ function createVoiceCommand() {
         };
         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);
@@ -1746,24 +1688,40 @@ function createVoiceCommand() {
     })
   );
   command.addCommand(
-    new Command9("create-custom").description("Design a custom voice and register it for self-hosted TTS (DashScope design \u2192 VoiceBox clone)").requiredOption("--prompt <text>", "Voice description (e.g. '\u5E74\u8F7B\u5973\u6027\uFF0C\u6E29\u6696\u4EB2\u5207\uFF0C\u6807\u51C6\u666E\u901A\u8BDD')").requiredOption("--preview-text <text>", "Sample text for voice preview").requiredOption("--name <name>", "Name for the custom voice").option("--language <lang>", "Language: zh, en, ja, ko, etc.", "zh").action(async function(options) {
+    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_prompt: options.prompt,
-          preview_text: options.previewText,
-          name: options.name,
-          language: options.language
+          voice: options.voice,
+          text: options.text
         };
-        const data = await ctx.client.post("/api/voice/create-custom", body);
-        printSuccess("voice.create_custom", data, ctx);
+        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_${Date.now()}.wav`);
+            writeFileSync5(outPath, buf);
+            data.local_path = outPath;
+          }
+        }
+        printSuccess("voice.synthesize", data, ctx);
       } catch (error2) {
-        printError("voice.create_custom", error2);
+        printError("voice.synthesize", error2);
       }
     })
   );
   command.addCommand(
-    new Command9("list").description("List custom cloned or designed voices").option("--type <type>", "Voice type: vc or vd").action(async function(options) {
+    new Command9("list").description("List custom designed voices").option("--type <type>", "Voice type: vd").action(async function(options) {
       try {
         const ctx = createContext(this);
         const data = await ctx.client.get(withVoiceType("/api/voice/list", options.type));
@@ -1774,7 +1732,7 @@ function createVoiceCommand() {
     })
   );
   command.addCommand(
-    new Command9("delete").description("Delete a custom cloned or designed voice").argument("<voice-id>", "Voice ID to delete").option("--type <type>", "Voice type: vc or vd").action(async function(voiceId, options) {
+    new Command9("delete").description("Delete a custom designed voice").argument("<voice-id>", "Voice ID to delete").option("--type <type>", "Voice type: vd").action(async function(voiceId, options) {
       try {
         const ctx = createContext(this);
         const path = withVoiceType(`/api/voice/${encodeURIComponent(voiceId)}`, options.type);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@doufunao123/asset-gateway",
-  "version": "0.14.3",
+  "version": "0.16.0",
   "description": "Universal asset generation gateway CLI",
   "type": "module",
   "bin": {