@kolbo/mcp 1.100.0 → 1.101.1

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@kolbo/mcp",
-  "version": "1.100.0",
+  "version": "1.101.1",
   "description": "Kolbo AI MCP Server - Generate images, videos, music, speech, and sound effects from Claude Code",
   "main": "src/index.js",
   "bin": {

package/skill/GENERATED.md

@@ -1,6 +1,6 @@
 # AUTO-GENERATED — do not edit
 
-This tree is mirrored from kolbo-code@e4764c5, the single source of truth.
+This tree is mirrored from kolbo-code@f1276f9, the single source of truth.
 Canonical source: packages/opencode/skills/kolbo/
 Distribution: .github/workflows/sync-skill-to-plugin.yml
 

package/src/apps/index.js

@@ -612,11 +612,19 @@ function sibling(models, hit, types) {
  * unchanged, so identifiers the catalog does not publish (hidden models) still
  * reach the API and it stays the source of truth.
  */
+const NATIVE_WORKFLOW_MODEL_IDS = new Set([
+  'kolbo-morphious-motion', 'kolbo-morphious-swap', 'kolbo-morphious-reframe',
+  'kolbo-morphious-lite-motion', 'kolbo-morphious-lite-swap',
+  'higgsfield-genjutsu-motion-transfer', 'higgsfield-genjutsu-object-swap',
+]);
 async function canonicalModelId(client, input, type) {
   if (!input || typeof input !== 'string') return input;
   const key = input.toLowerCase().trim();
   const want = normId(key);
-  if (!want || AUTO_ALIASES.has(want)) return input;
+  if (!want || AUTO_ALIASES.has(want)) return input;
+  // Exact native-workflow ids pass through: the near-miss check below would reject an
+  // unpublished or cached-out 'kolbo-*' id as unknown.
+  if (NATIVE_WORKFLOW_MODEL_IDS.has(key)) return input;
 
   let all;
   try {

package/src/tools/generate.js

@@ -233,6 +233,15 @@ const promptsField = (what) => z.array(z.string()).max(MAX_BATCH_PROMPTS).option
   `BATCH MODE — several DIFFERENT prompts (2–${MAX_BATCH_PROMPTS}) generated concurrently in ONE call and rendered together in ONE combined widget. **Hard cap: ${MAX_BATCH_PROMPTS} prompts per call — more than that is REJECTED with an error (never silently truncated), so split a longer list across several calls of at most ${MAX_BATCH_PROMPTS}.** Whenever the user wants multiple distinct ${what} with their own prompts, ALWAYS pass them all here instead of making several separate calls — separate calls clutter the chat with stacked widgets. All prompts share the same model/settings. When set, \`prompt\` is ignored. For N variations of a SINGLE prompt use num_images (image tools); for an AI-planned coherent scene set use generate_creative_director.`
 );
 
+// Native workflows: exactly one source video, prompt optional (the server analyzes the media).
+const MORPHIOUS_MODELS = new Set(['kolbo-morphious-motion', 'kolbo-morphious-swap', 'kolbo-morphious-reframe', 'kolbo-morphious-lite-motion', 'kolbo-morphious-lite-swap']);
+const PROMPTLESS_NATIVE = new Set([...MORPHIOUS_MODELS, 'higgsfield-genjutsu-motion-transfer', 'higgsfield-genjutsu-object-swap']);
+const MORPHIOUS_GUIDE = ' MORPHIOUS / GENJUTSU (exactly ONE source video, omit duration - it comes from the source): higgsfield-genjutsu-motion-transfer and higgsfield-genjutsu-object-swap accept an omitted prompt, and so do the kolbo-morphious-* models (Morphious only needs the source video and references; it analyzes them itself). '
+  + 'kolbo-morphious-motion 1-30 images, 4-30s; kolbo-morphious-swap 0-30 images, 4-30s; both 480p/720p/1080p, output keeps the source shape. '
+  + 'Lite (kolbo-morphious-lite-motion 1-10 images / kolbo-morphious-lite-swap 0-10 images) caps at 15s. '
+  + 'kolbo-morphious-reframe: NO images/DNA, 1-300s, aspect_ratio REQUIRED (16:9, 9:16, 1:1, 4:3, 3:4, 21:9), 540p or 720p. '
+  + 'higgsfield-genjutsu-motion-transfer / -object-swap: 1-8 images or Visual DNA, 4-30s, 480p/720p/1080p.';
+
 function registerGenerateTools(server, client, options = {}) {
   // Every JSON POST from these tools rehosts local file paths into the media
   // library first (see local-rehost.js) — the API only understands URLs.
@@ -1549,9 +1558,9 @@ function registerGenerateTools(server, client, options = {}) {
   // ─── generate_elements ─────────────────────────────────────
   server.tool(
     'generate_elements',
-    'Generate a video from reference elements (images, videos, and/or audio) + a text prompt. SPECIAL MODEL EXCEPTION: seedance-2-5-multilingual accepts an ordinary prompt in any language with the ORIGINAL target-language dialogue in double quotes, including Hebrew script. Use this tool even without references for that model, duration 8 or 12, resolution 480p/720p/1080p (no Draft), up to 3 still references or Visual DNAs, no video/audio references. It performs translation, distinct designed voices and segmented lip sync internally; never transliterate or pre-translate its quoted dialogue. The following regular Seedance instructions apply only to other model IDs. Use when the user wants to animate specific uploaded/referenced assets — e.g. "animate this product", "put these 3 characters into a scene". PRIMARY ROUTE FOR A DNA-ANCHORED MULTI-SHOT FILM: one call can carry the whole sequence — seedance-2-5 takes 4-30s, up to 30 shots and 20 Visual DNAs in a SINGLE generation (seedance-2: 4-15s, 9 DNAs) — instead of a stack of separate clips. DIALOGUE IS PERFORMED NATIVELY: quoted dialogue in the prompt comes back as synced voices with lip movement, room tone and the SFX named in the AUDIO block — never route scene dialogue to generate_speech or generate_lipsync. Write dialogue in ENGLISH or Latin transliteration of Hebrew ("shalom"), never Hebrew script — Seedance does not speak Hebrew; prefer Gemini Omni Flash 1.1 or Gemini Omni 1 for native Hebrew. COST: resolution is a multiplier. When list_models publishes `video_input_credit` and this call carries videos, charge that rate against nominal input seconds + nominal output seconds; MP4 padding within 0.15s of an integer snaps to that integer and larger fractions round up. Otherwise use the normal output-second rate. PROMPT CONTRACT (Seedance / Elements): Locked Intro only — Total line, then [GLOBAL LOOK] / [CAST] / [LOCATION] / SHOT N. Do NOT write SCENE CONTEXT / OPTICS / ACTION department packs. Every Visual DNA in visual_dna_ids MUST also appear in the prompt as @ExactDNAName (e.g. "@Zohar walks…") — never "Zohar\'s" or "the man on the left" as a substitute. IMPORTANT: different models accept different numbers and durations of inputs — call list_models type="elements" and read elements_max_images / elements_max_videos / elements_max_audio plus min_video_duration / max_video_duration before generating. For text-only → video use generate_video instead. For animating a single still image use generate_video_from_image. Returns the final video URL when complete.',
+    'Generate a video from reference elements (images, videos, and/or audio) + a text prompt. SPECIAL MODEL EXCEPTION: seedance-2-5-multilingual accepts an ordinary prompt in any language with the ORIGINAL target-language dialogue in double quotes, including Hebrew script. Use this tool even without references for that model, duration 8 or 12, resolution 480p/720p/1080p (no Draft), up to 3 still references or Visual DNAs, no video/audio references. It performs translation, distinct designed voices and segmented lip sync internally; never transliterate or pre-translate its quoted dialogue. The following regular Seedance instructions apply only to other model IDs. Use when the user wants to animate specific uploaded/referenced assets — e.g. "animate this product", "put these 3 characters into a scene". PRIMARY ROUTE FOR A DNA-ANCHORED MULTI-SHOT FILM: one call can carry the whole sequence — seedance-2-5 takes 4-30s, up to 30 shots and 20 Visual DNAs in a SINGLE generation (seedance-2: 4-15s, 9 DNAs) — instead of a stack of separate clips. DIALOGUE IS PERFORMED NATIVELY: quoted dialogue in the prompt comes back as synced voices with lip movement, room tone and the SFX named in the AUDIO block — never route scene dialogue to generate_speech or generate_lipsync. Write dialogue in ENGLISH or Latin transliteration of Hebrew ("shalom"), never Hebrew script — Seedance does not speak Hebrew; prefer Gemini Omni Flash 1.1 or Gemini Omni 1 for native Hebrew. COST: resolution is a multiplier. When list_models publishes `video_input_credit` and this call carries videos, charge that rate against nominal input seconds + nominal output seconds; MP4 padding within 0.15s of an integer snaps to that integer and larger fractions round up. Otherwise use the normal output-second rate. PROMPT CONTRACT (Seedance / Elements): Locked Intro only — Total line, then [GLOBAL LOOK] / [CAST] / [LOCATION] / SHOT N. Do NOT write SCENE CONTEXT / OPTICS / ACTION department packs. Every Visual DNA in visual_dna_ids MUST also appear in the prompt as @ExactDNAName (e.g. "@Zohar walks…") — never "Zohar\'s" or "the man on the left" as a substitute. IMPORTANT: different models accept different numbers and durations of inputs — call list_models type="elements" and read elements_max_images / elements_max_videos / elements_max_audio plus min_video_duration / max_video_duration before generating. For text-only → video use generate_video instead. For animating a single still image use generate_video_from_image. Returns the final video URL when complete.' + MORPHIOUS_GUIDE,
     {
-      prompt: z.string().describe('For seedance-2-5-multilingual: ordinary scene description with original target-language dialogue in double quotes; no manual translation or transliteration. For other models, Locked Intro prompt (Seedance/Elements): Total line, [GLOBAL LOOK], [CAST] with @ExactDNAName for every visual_dna_ids entry, [LOCATION], then SHOT N. Not SCENE CONTEXT/OPTICS/ACTION packs. Never substitute "the left man" or "Zohar\'s" for @Name. EVERY attached reference must also be tagged by its 1-based array position — `@Image 1`/`@Image 2` (reference_images), `@Video 1` (reference_videos), `@Audio 1` (reference_audio_urls) — and its job stated ("@Image 1 defines the character\'s face and wardrobe", "@Video 1 defines the camera move"). An untagged attachment is ignored by the engine even though it was uploaded and billed.'),
+      prompt: z.string().optional().describe('Optional ONLY for kolbo-morphious-* and higgsfield-genjutsu-* models (the server analyzes the media). For seedance-2-5-multilingual: ordinary scene description with original target-language dialogue in double quotes; no manual translation or transliteration. For other models, Locked Intro prompt (Seedance/Elements): Total line, [GLOBAL LOOK], [CAST] with @ExactDNAName for every visual_dna_ids entry, [LOCATION], then SHOT N. Not SCENE CONTEXT/OPTICS/ACTION packs. Never substitute "the left man" or "Zohar\'s" for @Name. EVERY attached reference must also be tagged by its 1-based array position — `@Image 1`/`@Image 2` (reference_images), `@Video 1` (reference_videos), `@Audio 1` (reference_audio_urls) — and its job stated ("@Image 1 defines the character\'s face and wardrobe", "@Video 1 defines the camera move"). An untagged attachment is ignored by the engine even though it was uploaded and billed.'),
       model: z.string().optional().describe('Model identifier. If the user already named a family (Grok / Kling / Veo / Seedance / …), pass THAT family — never default to Seedance because Elements often uses it. Use list_models type="elements" for exact ids and elements_max_* caps. Do NOT omit (omitting = Smart Select).'),
       reference_images: z.array(z.string()).optional().describe('Array of image references (product shots, character references, etc.). Tag each one in the prompt text by its 1-based position here — item 1 is `@Image 1`, item 2 is `@Image 2` — or the engine ignores it. Accepts a public URL (forwarded as-is; if the API rejects an external URL as untrusted, it is auto-rehosted into the media library and retried once) OR an absolute local path, which is uploaded for you. **Cap: pass at most `elements_max_images` URLs from list_models for the chosen model — exceeding it is a deterministic 400.**'),
       reference_videos: z.array(z.string()).optional().describe('Array of reference videos for models that accept video inputs. Tag each one in the prompt text by its 1-based position here — `@Video 1`, `@Video 2` — stating what it defines (motion, camera move, pacing), or the engine ignores it. Accepts a public URL (forwarded as-is; if the API rejects an external URL as untrusted, it is auto-rehosted into the media library and retried once) OR an absolute local path, which is uploaded for you. **Cap: pass at most `elements_max_videos` URLs and keep every clip within `min_video_duration`-`max_video_duration` from list_models.** If `video_input_credit` is present, every attached video contributes its nominal duration to combined-second billing: encoder padding within 0.15s of an integer snaps to it; larger fractions round up.'),
@@ -1577,14 +1586,14 @@ function registerGenerateTools(server, client, options = {}) {
       project_id: projectIdField,
       session_id: sessionIdField
     },
-    async ({ prompt, model, reference_images, reference_videos, reference_audio_urls, audio_url, files, duration, aspect_ratio, motion, preset_id, enhance_prompt = false, visual_dna_ids, resolution, draft, sound_enabled, keyframes, multi_shots, multi_shot_count, session_name, project_id, session_id }) => {
+    async ({ prompt = '', model, reference_images, reference_videos, reference_audio_urls, audio_url, files, duration, aspect_ratio, motion, preset_id, enhance_prompt = false, visual_dna_ids, resolution, draft, sound_enabled, keyframes, multi_shots, multi_shot_count, session_name, project_id, session_id }) => {
       validateVideoPrompt({ prompt, duration, aspect_ratio, multi_shots, multi_shot_count });
       if (draft === true) resolution = '480p-draft';
       else if (draft === false && resolution?.endsWith('-draft')) resolution = resolution.slice(0, -6);
       model = await canonicalModelId(client, model, 'elements'); // lenient id resolution ("z-image" → "z-image/turbo")
       aspect_ratio = await resolveCatalogAspectRatio(client, model, aspect_ratio, 'elements');
       validateVideoPrompt({ prompt, duration, aspect_ratio, multi_shots, multi_shot_count });
-      if (!prompt) throw new Error('prompt is required');
+      if (!prompt && !PROMPTLESS_NATIVE.has(model)) throw new Error('prompt is required');
 
       // Elements is the one tool that takes all three modalities, and either a
       // URL or a local path for any of them. Bucket every input by the kind the
@@ -1970,10 +1979,10 @@ function registerGenerateTools(server, client, options = {}) {
   // ─── generate_video_from_video ─────────────────────────────
   server.tool(
     'generate_video_from_video',
-    'Restyle / transform an existing video (video-to-video). Use for style transfer, scene restyling, subject swap, motion transfer, character replacement, or burning in styled subtitles (VEED Subtitles). Source video can be a URL or absolute local path. `prompt` is OPTIONAL: most models need it, but prompt-less models (VEED Subtitles, Act Two, Wan Animate, Kling Motion Control) ignore it. For VEED Subtitles, pass a `preset` style and optional `source_language` / `translation_language` instead of a prompt. IMPORTANT: different models support different extra inputs — call list_models type="video_to_video" and read max_images / max_videos / max_elements on the chosen model before generating. Pass reference_images for models with max_images > 0 (e.g. Kling O1/O3, Aleph, WAN VACE), reference_videos for models with max_videos > 1 (e.g. WAN 2.6 reference-to-video accepts up to 3), and elements for models with max_elements > 0. REPAIR / RETIME (not restyling, no prompt needed — these keep the footage and fix or retime it): model "topaz/deblur/video" removes lens, motion and compression blur; "topaz/colorize/video" colorizes black-and-white footage; "topaz/interpolate/video" is SLOW MOTION and frame-rate conversion (see slowdown_factor / target_fps); "topaz/sdr-to-hdr/video" masters SDR footage to HDR (see output_format). Reach for these when the user says blurry, shaky-detail, black-and-white, slow motion, smoother frame rate, or HDR — a restyle model would repaint the video instead of repairing it. For animating a still image use generate_video_from_image instead. For text-only → video use generate_video.',
+    'Restyle / transform an existing video (video-to-video). Use for style transfer, scene restyling, subject swap, motion transfer, character replacement, or burning in styled subtitles (VEED Subtitles). Source video can be a URL or absolute local path. `prompt` is OPTIONAL: most models need it, but prompt-less models (VEED Subtitles, Act Two, Wan Animate, Kling Motion Control) ignore it. For VEED Subtitles, pass a `preset` style and optional `source_language` / `translation_language` instead of a prompt. IMPORTANT: different models support different extra inputs — call list_models type="video_to_video" and read max_images / max_videos / max_elements on the chosen model before generating. Pass reference_images for models with max_images > 0 (e.g. Kling O1/O3, Aleph, WAN VACE), reference_videos for models with max_videos > 1 (e.g. WAN 2.6 reference-to-video accepts up to 3), and elements for models with max_elements > 0. REPAIR / RETIME (not restyling, no prompt needed — these keep the footage and fix or retime it): model "topaz/deblur/video" removes lens, motion and compression blur; "topaz/colorize/video" colorizes black-and-white footage; "topaz/interpolate/video" is SLOW MOTION and frame-rate conversion (see slowdown_factor / target_fps); "topaz/sdr-to-hdr/video" masters SDR footage to HDR (see output_format). Reach for these when the user says blurry, shaky-detail, black-and-white, slow motion, smoother frame rate, or HDR — a restyle model would repaint the video instead of repairing it. For animating a still image use generate_video_from_image instead. For text-only → video use generate_video.' + MORPHIOUS_GUIDE,
     {
       source_video: z.string().describe('URL or absolute local path to the primary source video to restyle. **Source duration must fall within `min_video_duration`-`max_video_duration` from list_models for the chosen model** — videos outside that range are rejected (or silently truncated by some upstream providers). For models that use reference_videos as their primary input (e.g. WAN 2.6 reference-to-video), pass the first reference video here and also include it in reference_videos.'),
-      prompt: z.string().optional().describe('Text description of the desired restyle / transformation. Required by most video-to-video models; omit for prompt-less models (VEED Subtitles, Act Two, Wan Animate, Kling Motion Control).'),
+      prompt: z.string().optional().describe('Text description of the desired restyle / transformation. Required by most video-to-video models; omit for prompt-less models (VEED Subtitles, Act Two, Wan Animate, Kling Motion Control, kolbo-morphious-*, higgsfield-genjutsu-motion-transfer, higgsfield-genjutsu-object-swap).'),
       model: z.string().optional().describe('Model identifier. Use list_models type="video_to_video" to see options and check max_images / max_videos / max_elements / max_video_duration per model. Pick a SPECIFIC model — do NOT omit (omitting = Smart Select auto-pick, which we avoid); call list_models for this type and choose the model that best fits the user\'s intent.'),
       aspect_ratio: z.string().optional().describe(aspectRatioDescribe() + ' Default: matches source.'),
       duration: z.number().optional().describe('Output duration in seconds. Must be in `supported_durations` from list_models, OR within `min_output_duration`-`max_output_duration`. Default: matches source'),

package/src/tools/video-prompt.js

@@ -1,6 +1,6 @@
 // Validate only explicit, machine-readable declarations. Never infer creative
 // intent from adjectives, require a template, or rewrite a user's prompt.
-function validateVideoPrompt({ prompt, duration, aspect_ratio, multi_shots, multi_shot_count }) {
+function validateVideoPrompt({ prompt = '', duration, aspect_ratio, multi_shots, multi_shot_count }) {
   const totals = [...prompt.matchAll(/^\s*Total:\s*(\d+(?:\.\d+)?)s\s*\/\s*(\d+)\s*shots?\s*\/\s*(\d+:\d+)\s*$/gim)]
     .map((m) => ({ duration: Number(m[1]), count: Number(m[2]), aspect: m[3] }));
   if (!totals.length) return; // Existing free-form clients remain supported.