@kolbo/mcp 1.113.2 → 1.114.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@kolbo/mcp",
-  "version": "1.113.2",
+  "version": "1.114.0",
   "description": "Kolbo AI MCP Server - Generate images, videos, music, speech, and sound effects from Claude Code",
   "main": "src/index.js",
   "bin": {

package/src/tools/generate.js

@@ -1950,6 +1950,18 @@ function registerGenerateTools(server, client, options = {}) {
       if (!source) throw new Error('source is required (URL or absolute local path to image/video)');
       if (!audio) throw new Error('audio is required (URL or absolute local path to audio file)');
 
+      // Upload local files first so the card can show the face source and the
+      // audio from the same CDN URLs the API receives (multipart left the card
+      // with no references at all).
+      ({ source, audio } = await rehostLocalPaths(client, { source, audio, project_id }, { allowLocalFiles: !options.remote }));
+      const isVideoUrl = (u) => /\.(mp4|mov|webm|mkv|avi|m4v)(\?|$)/i.test(u);
+      const httpRef = (u) => (typeof u === 'string' && /^https?:\/\//i.test(u) ? [u] : []);
+      const lipsyncRefs = {
+        reference_images: httpRef(source).filter((u) => !isVideoUrl(u)),
+        reference_videos: httpRef(source).filter(isVideoUrl),
+        reference_audio: httpRef(audio),
+      };
+
       const sourceIsUrl = typeof source === 'string' && /^https?:\/\//i.test(source);
       const audioIsUrl = typeof audio === 'string' && /^https?:\/\//i.test(audio);
 
@@ -2005,8 +2017,7 @@ function registerGenerateTools(server, client, options = {}) {
 
       if (returnsImmediately()) return submittedResult({
         tool: 'generate_lipsync', kind: 'video', gen: startResponse, client, model,
-        prompt: text_prompt, settings: { mode: 'lipsync' },
-        reference_images: sourceIsUrl && !/\.(mp4|mov|webm|mkv|avi|m4v)(\?|$)/i.test(source) ? [source] : [],
+        prompt: text_prompt, settings: { mode: 'lipsync' }, ...lipsyncRefs,
       });
 
       const poll = await pollOrTimedOut(client, startResponse.generation_id, {
@@ -2018,9 +2029,7 @@ function registerGenerateTools(server, client, options = {}) {
 
       return uiCompleted({
         tool: 'generate_lipsync', kind: 'video', gen: startResponse, client, model,
-        prompt: text_prompt, settings: { mode: 'lipsync' },
-        reference_images: sourceIsUrl && !/\.(mp4|mov|webm|mkv|avi|m4v)(\?|$)/i.test(source) ? [source] : [],
-        reference_videos: sourceIsUrl && /\.(mp4|mov|webm|mkv|avi|m4v)(\?|$)/i.test(source) ? [source] : [],
+        prompt: text_prompt, settings: { mode: 'lipsync' }, ...lipsyncRefs,
         urls: result.result?.urls || [],
         thumbnail_url: result.result?.thumbnail_url || null,
         duration: result.result?.duration || null,

package/src/tools/voices.js

@@ -113,11 +113,11 @@ function registerVoiceTools(server, client, options = {}) {
   // ─── clone_voice ───────────────────────────────────────────
   server.tool(
     'clone_voice',
-    'Clone a custom TTS voice from an audio sample (the user\'s voice or any voice they own the rights to). Costs credits (provider-dependent) — confirm with the user before firing. Providers: "elevenlabs" (recommended, 2s–180s sample) or "deepdub" (2s–300s). After cloning, the voice appears in `list_voices` and can be used with generate_speech. Free plan caps at 3 custom voices.',
+    'Clone a custom TTS voice from an audio sample (the user\'s voice or any voice they own the rights to). One-time setup cost depends on the provider — confirm with the user before firing. Providers (sample length, setup credits): "elevenlabs" (recommended, 8s–180s, 30), "kolbo" (Kolbo Voice Clone, 1s–60s, free), "fish-audio" (10s–120s, free), "cartesia" (3s–60s, 10), "heygen" (10s–180s, 75), "deepdub" (2s–300s, free), "minimax" (10s–180s, 400). After cloning, the voice appears in `list_voices` and can be used with generate_speech. Free plan caps at 3 custom voices.',
     {
       audio: z.string().describe('URL or absolute local path of the voice sample audio (any common format; converted server-side).'),
       voice_name: z.string().describe('Name for the cloned voice.'),
-      provider: z.enum(['elevenlabs', 'deepdub']).optional().describe('Cloning provider. Default: elevenlabs (recommended).'),
+      provider: z.enum(['elevenlabs', 'kolbo', 'fish-audio', 'cartesia', 'heygen', 'deepdub', 'minimax']).optional().describe('Cloning provider (same set as the app\'s voice cloning). Default: elevenlabs (recommended).'),
       language: z.string().optional().describe('Optional language hint (auto-detected when omitted).'),
       project_id: z.string().optional().describe('Project to associate the voice with (from list_projects). Omit for the default project.')
     },
@@ -126,7 +126,8 @@ function registerVoiceTools(server, client, options = {}) {
       const form = new FormData();
       form.append('audioFile', resolved.buffer, { filename: resolved.filename, contentType: resolved.contentType });
       form.append('voiceName', voice_name);
-      form.append('provider', provider || 'elevenlabs');
+      // "kolbo" is the user-facing name of Kolbo Voice Clone; the API keeps its internal id.
+      form.append('provider', provider === 'kolbo' ? 'zonos' : (provider || 'elevenlabs'));
       if (language) form.append('language', language);
       if (project_id) form.append('project_id', project_id);
       const result = await client.postMultipart('/v1/voices/clone', form);