@kolbo/mcp 1.112.2 → 1.112.5

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@kolbo/mcp",
-  "version": "1.112.2",
+  "version": "1.112.5",
   "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@66aa15c, the single source of truth.
+This tree is mirrored from kolbo-code@49cd478, the single source of truth.
 Canonical source: packages/opencode/skills/kolbo/
 Distribution: .github/workflows/sync-skill-to-plugin.yml
 

package/skill/SKILL.md

@@ -1,5 +1,5 @@
 ---
-version: 0.9.22
+version: 0.9.23
 name: kolbo
 description: |
   Generate, edit, analyze, and direct creative media through Kolbo AI: images,
@@ -155,7 +155,7 @@ Font tools (when exposed by the installed MCP): `list_fonts`, `get_font`, `uploa
 | `list_models` / `list_voices` / `check_credits` / `show_plans` / `get_generation_status` / `cancel_generation` / `get_session_usage` | Discovery + status. `list_models` with no args returns the recommended shortlist out of ~428 — pass `type` for a full category with per-model caps. `cancel_generation` stops an in-flight job and refunds what it can: use it when the user changes their mind mid-generation instead of letting it run. `show_plans` renders the balance + upgrade card for pricing/plan/upgrade questions. |
 | `download_media_from_url` / `get_download_status` / `cancel_download` | Download a public social/video page URL to Kolbo storage. Async job; use download status, not generation status. See `workflows/media-library.md`. |
 | `upload_media` / `create_upload_ticket` / `list_media` / `get_media` / `get_media_stats` / `favorite_media` / `unfavorite_media` / `delete_media` / `restore_media` / `permanently_delete_media` / `move_media` / `bulk_*_media` / `*_media_folder` | Media library — see `workflows/media-library.md`. Getting a LOCAL file in depends on where the server runs: `upload_media` with a path only works on a local (stdio) install; over a remote connector use `create_upload_ticket` and POST the file yourself. |
-| `create_visual_dna` / `update_visual_dna` / `generate_character_sheet` / `list_visual_dnas` / `get_visual_dna` / `delete_visual_dna` / `*_visual_dna_folder` (5 folder tools) | Visual DNA (+ character sheet, character folders) — see `workflows/visual-dna.md`. Edit with `update_visual_dna`; never delete+recreate. |
+| `create_visual_dna` / `update_visual_dna` / `generate_character_sheet` / `list_visual_dnas` / `get_visual_dna` / `delete_visual_dna` / `*_visual_dna_folder` (5 folder tools) | Visual DNA (+ character sheet and character folders) — see `workflows/visual-dna.md`. Edit with `update_visual_dna`; never delete+recreate. Account transfers use the Kolbo UI until the MCP tools ship. |
 | `list_moodboards` / `get_moodboard` / `list_presets` / `list_cinematic_presets` | Style overlays + presets. `list_presets` spans FOUR distinct catalogs (`image`, `image_edit`, `video`, `music`; `text_to_video` is an alias for `video`, `shorts` is empty) — the `video` one holds 200+ Seedance shot recipes. `list_cinematic_presets` is a separate tool feeding the `cinematic` arg, never `preset_id`. Full doctrine + intent→catalog map: `references/workflows/presets.md`. Never omit `preset_id` after claiming a preset was used. |
 | `list_color_palettes` / `analyze_color_palette` / `create_color_palette` / `update_color_palette` / `delete_color_palette` / `activate_color_palette` / `deactivate_color_palette` | **Color DNA — sticky + account-wide; at most one palette active at a time**, and while active it strict-grades **every** image and video generation automatically. Per-generation opt-out: `skip_color_palette: true`. Details: `workflows/color-dna.md`. |
 | `list_agents` / `create_agent` / `update_agent` / `delete_agent` (same tools as `list_skills` / `create_skill` / `update_skill` / `delete_skill`) | Custom chat agents — reusable named personas for `chat_send_message`. The agent's `description` IS the system instruction. Resolve a name the user mentions ("use my SEO agent") to an id with `list_agents` (paged: `page` / `limit`, follow `pagination.has_more`), then pass `agent_id`. Global/preset agents are read-only; only the user's own can be updated or deleted. |

package/skill/VERSION

@@ -1 +1 @@
-0.9.22
+0.9.23

package/skill/references/workflows/presets.md

@@ -1,5 +1,11 @@
 # Presets — Kolbo's stored instruction blocks
 
+Generation widgets show the selected template name and thumbnail in a dedicated
+Template section, plus every supplied reference in References. Local image sources
+preview through their uploaded CDN URLs. Both sections persist through completion,
+including image batches and standalone status checks. Keep passing the exact
+`preset_id`; the thumbnail is a preview, not an extra conditioning image.
+
 A preset is a long, hand-tuned instruction block stored in Kolbo's catalog. Passing
 `preset_id` prepends it to the user's prompt at generation time. **The craft lives on
 the preset** — a three-panel character sheet, a Seedance shot-sequence structure, a

package/skill/references/workflows/visual-dna.md

@@ -254,6 +254,18 @@ Tools: `list_visual_dna_folders`, `create_visual_dna_folder` (`name`, optional h
 - **Creating many characters for one production?** Create the folder FIRST, then `move_visual_dna_to_folder` each DNA right after `create_visual_dna` — don't leave a big cast unsorted at root.
 - To list a folder's contents: `list_visual_dnas` and filter by each profile's `folder_id` (there is no server-side folder filter).
 
+## Transfer through Kolbo UI — send DNAs to another account
+
+Use the authenticated Kolbo UI for account transfers. Dedicated MCP transfer tools
+are not in the current published tool surface; do not invent or call them.
+
+- Recipient must already have a Kolbo account; nothing changes until they accept (7-day expiry).
+- On accept the recipient gets their OWN copy — **new ids**, returned in `copies`. Use those ids and names for the recipient's generations, never the sender's ids. A name clash becomes "Name (1)".
+- `keep_sender_copy: false` is a move: the sender's originals go to trash on accept. Default keeps them.
+- Only the sender's personal, ready DNAs — not global, organization, or still-generating ones. Custom cloned/designed voices stay owned by the sender.
+- GATE: confirm recipient email, the DNA list and keep/move with the user before sending through the UI; confirm before accepting.
+- Whole project instead (cast comes along as copies)? Use `transfer_project`.
+
 ## Character sheet — default for production assets (not a catalog preset)
 
 Custom instructions live on the **image preset**. Resolve it silently, then generate:

package/src/apps/theme.js

@@ -152,7 +152,16 @@ body {
   align-items: center; justify-content: center; font-family: inherit;
 }
 .k-chip img.k-voice-thumb { width: 18px; height: 18px; border-radius: 999px; margin-left: -3px; }
-.k-ref-thumb { width: 26px; height: 26px; border-radius: 6px; object-fit: cover; border: 1px solid var(--border-strong); }
+.k-ref-thumb { width: 26px; height: 26px; border-radius: 6px; object-fit: cover; border: 1px solid var(--border-strong); }
+.k-input-section { flex-basis: 100%; min-width: 0; margin-top: 6px; }
+.k-input-label { display: block; color: var(--muted); font-size: 11px; margin-bottom: 6px; }
+.k-input-media { display: flex; flex-wrap: wrap; align-items: center; gap: 8px; }
+.k-input-media .k-ref-thumb { width: 40px; height: 40px; object-fit: contain; background: var(--panel); }
+.k-ref-audio { display: flex; flex-direction: column; gap: 6px; max-width: 100%; font-size: 11px; }
+.k-ref-audio audio { width: 240px; max-width: 100%; height: 32px; }
+.k-template { display: flex; align-items: center; gap: 12px; font-size: 12px; }
+.k-template-thumb, .k-template-placeholder { width: 104px; height: 80px; border-radius: 8px; object-fit: contain; background: var(--panel); border: 1px solid var(--border-strong); }
+.k-template-placeholder { display: flex; align-items: center; justify-content: center; color: var(--muted); }
 .k-peek-hit { cursor: zoom-in; }
 .k-peek {
   position: absolute; inset: 0; z-index: 20;

package/src/apps/widgets/generation.js

@@ -385,7 +385,7 @@ function renderChips(sc) {
   if (s.mode) h += chip(esc(editLabel(s.mode)));
   if (Array.isArray(s.details)) s.details.forEach(function (d) { if (d) h += chip(esc(String(d))); });
   if (many && !kind) h += chip(sc.count + ' items');
-  h += referenceHTML(sc);
+  h += templateHTML(sc) + referenceHTML(sc);
   el('chips').innerHTML = h;
 }
 
@@ -429,7 +429,7 @@ function referenceHTML(sc) {
       h += '<video class="k-ref-thumb k-peek-hit" src="' + url + '#t=0.1" muted playsinline preload="metadata" title="'
         + title + '"' + peek + ' onerror="this.style.display=\\'none\\'"></video>';
     } else if (refs[i].kind === 'audio') {
-      h += '<span class="k-chip" title="' + title + '">' + ICONS.sound + ' audio ref</span>';
+      h += '<span class="k-ref-audio" title="' + title + '">' + ICONS.sound + ' audio ref<audio controls preload="none" src="' + url + '"></audio></span>';
     } else {
       h += '<img class="k-ref-thumb k-peek-hit" src="' + url + '" alt="" loading="lazy" title="' + title + '"'
         + peek + ' onerror="this.style.display=\\'none\\'">';
@@ -438,12 +438,23 @@ function referenceHTML(sc) {
   // Local files the caller passed by path: the iframe cannot load them and they
   // are only uploaded during the run, so name them instead of showing nothing.
   // The finished card carries the uploaded thumbnails and drops these.
-  if (sc.phase !== 'completed' && Array.isArray(sc.local_references)) {
-    sc.local_references.slice(0, 6).forEach(function (name) {
+  if (sc.phase !== 'completed' && Array.isArray(sc.local_references) && !refs.length) {
+    sc.local_references.forEach(function (name) {
       h += '<span class="k-chip" title="' + esc(name) + '">' + ICONS.image + ' ' + esc(String(name).slice(0, 24)) + '</span>';
     });
   }
-  return h;
+  return h ? '<div class="k-input-section"><span class="k-input-label">References</span><div class="k-input-media">' + h + '</div></div>' : '';
+}
+function templateHTML(sc) {
+  var s = sc.settings || {};
+  if (!s.preset_id && !s.preset_name && !s.preset_thumbnail) return '';
+  var name = s.preset_name || 'Selected template';
+  var url = s.preset_thumbnail;
+  var media = url ? '<img class="k-template-thumb k-peek-hit" src="' + esc(url) + '" alt="' + esc(name)
+    + '" loading="lazy" data-peek="' + esc(url) + '" data-peek-kind="image" data-peek-cap="' + esc(name) + '">'
+    : '<span class="k-template-placeholder">' + ICONS.image + '</span>';
+  return '<div class="k-input-section"><span class="k-input-label">Template</span><div class="k-template">'
+    + media + '<span>' + esc(name) + '</span></div></div>';
 }
 // Which characters/looks are locked into this generation — by face and name,
 // resolved from visual_dna_ids server-side. "1 Visual DNA" told the user nothing
@@ -830,6 +841,7 @@ function poll(sc) {
 
       var done = Object.assign({}, sc, r, {
         phase: 'completed',
+        settings: Object.assign({}, sc.settings || {}, r.settings || {}),
         urls: r.urls || st.urls || [],
         credits_used: st.credits_used != null ? st.credits_used : sc.credits_used,
         // Status structuredContent used to set open_url:undefined and wipe the
@@ -1462,7 +1474,7 @@ function kindFromTool(tool, sc) {
 function liveFromTimedOut(sc) {
   if (!sc || !sc.generation_id || isTerminal(sc.state) || Array.isArray(sc.urls)) return null;
   var ids = Array.isArray(sc.generation_ids) && sc.generation_ids.length > 1 ? sc.generation_ids : null;
-  var live = {
+  var live = Object.assign({}, preRefSc(originTool, originArgs), {
     widget: 'generation',
     phase: 'generating',
     tool: originTool,
@@ -1477,7 +1489,8 @@ function liveFromTimedOut(sc) {
       quality: originArgs.quality,
       mode: originArgs.operation,
       visual_dna_ids: originArgs.visual_dna_ids,
-      moodboard_id: originArgs.moodboard_id
+      moodboard_id: originArgs.moodboard_id,
+      preset_id: originArgs.preset_id
     }, originArgs),
     count: ids ? ids.length : (originArgs.num_images || 1),
     generation_id: sc.generation_id,
@@ -1485,14 +1498,14 @@ function liveFromTimedOut(sc) {
     status_args: ids ? { generation_ids: ids, wait: true } : { generation_id: sc.generation_id, wait: true },
     session_id: sc.session_id || originArgs.session_id,
     project_id: sc.project_id || originArgs.project_id
-  };
+  });
   if (ids) { live.generation_ids = ids; live.prompts = originArgs.prompts || []; }
   return live;
 }
 
 function completedFromPlain(sc) {
   if (!sc || (!Array.isArray(sc.urls) && !Array.isArray(sc.scenes))) return null;
-  return Object.assign({}, sc, {
+  return Object.assign({}, preRefSc(originTool, originArgs), sc, {
     widget: 'generation',
     phase: 'completed',
     tool: originTool,
@@ -1505,7 +1518,8 @@ function completedFromPlain(sc) {
       resolution: originArgs.draft === true ? '480p-draft' : originArgs.draft === false ? String(originArgs.resolution || '').replace(/-draft$/i, '') : originArgs.resolution,
       aspect_ratio: originArgs.aspect_ratio,
       quality: originArgs.quality,
-      mode: originArgs.operation
+      mode: originArgs.operation,
+      preset_id: originArgs.preset_id
     }, originArgs),
     urls: sc.urls || [],
     session_id: sc.session_id || originArgs.session_id,

package/src/tools/generate.js

@@ -110,7 +110,7 @@ async function submitBatch(rawItems, submitOne) {
 // multi-id branch, same fix: always ship structuredContent via the shared
 // kind:'status' grid, which already renders any mix of completed/processing
 // items correctly.
-async function pollBatch(client, batch, { interval, timeout }, toolName, submittedModel) {
+async function pollBatch(client, batch, { interval, timeout }, toolName, submittedModel, inputs = {}) {
   const polls = await Promise.all(batch.ids.map((id) => pollOrTimedOut(client, id, { interval, timeout })));
   const generations = polls.map((p, i) => p.timedOut
     ? { prompt: batch.ok[i].prompt, generation_id: batch.ids[i], status: 'processing', note: 'Still running — call get_generation_status with wait=true to collect it.' }
@@ -126,8 +126,9 @@ async function pollBatch(client, batch, { interval, timeout }, toolName, submitt
   return uiCompleted({
     tool: toolName, kind: 'status', client,
     model: modelsRan.length === 1 ? modelsRan[0] : (submittedModel || 'Generations'),
-    gen: { generation_id: batch.ids[0], session_id: batch.ok[0].gen.session_id },
-    settings: {},
+    gen: { generation_id: batch.ids[0], session_id: batch.ok[0].gen.session_id },
+    settings: {},
+    ...inputs,
     items: generations.map(g => ({
       id: g.generation_id,
       state: g.status,
@@ -338,10 +339,11 @@ function registerGenerateTools(server, client, options = {}) {
       if (!prompt && !(prompts && prompts.length)) throw new Error('Provide prompt or prompts');
       model = await canonicalModelId(client, model, 'text_to_img'); // lenient id resolution ("z-image" → "z-image/turbo")
       aspect_ratio = await resolveCatalogAspectRatio(client, model, aspect_ratio, 'text_to_img');
-      const shared = {
-        model, aspect_ratio, enhance_prompt, font_ids,
-        reference_images, visual_dna_ids, moodboard_id, enable_web_search, resolution, quality, background, output_format, output_compression, moderation, mask_image_url, preset_id, cinematic, skip_color_palette, project_id, session_id
-      };
+      const shared = await rehostLocalPaths(client, {
+        model, aspect_ratio, enhance_prompt, font_ids,
+        reference_images, visual_dna_ids, moodboard_id, enable_web_search, resolution, quality, background, output_format, output_compression, moderation, mask_image_url, preset_id, cinematic, skip_color_palette, project_id, session_id
+      }, { allowLocalFiles: !options.remote });
+      reference_images = [...(shared.reference_images || []), shared.mask_image_url].filter(Boolean);
 
       // Batch mode: N different prompts, one widget owning all generation ids.
       if (prompts && prompts.length) {
@@ -354,7 +356,7 @@ function registerGenerateTools(server, client, options = {}) {
           status_args: { generation_ids: batch.ids, wait: true },
           reference_images
         });
-        return pollBatch(client, batch, { interval: (batch.ok[0].gen.poll_interval_hint || 3) * 1000, timeout: 150000 }, 'generate_image', model);
+        return pollBatch(client, batch, { interval: (batch.ok[0].gen.poll_interval_hint || 3) * 1000, timeout: 150000 }, 'generate_image', model, { settings: imageSettings(shared), reference_images });
       }
 
       const gen = await client.post('/v1/generate/image', { ...shared, prompt, num_images });
@@ -429,10 +431,12 @@ function registerGenerateTools(server, client, options = {}) {
       if (!prompt && !(prompts && prompts.length)) throw new Error('Provide prompt or prompts');
       model = await canonicalModelId(client, model, 'image_editing'); // lenient id resolution ("z-image" → "z-image/turbo")
       aspect_ratio = await resolveCatalogAspectRatio(client, model, aspect_ratio, 'image_editing');
-      const shared = {
-        model, source_images, reference_images, aspect_ratio, enhance_prompt, font_ids,
-        visual_dna_ids, moodboard_id, enable_web_search, resolution, quality, background, output_format, output_compression, moderation, mask_image_url, preset_id, cinematic, skip_color_palette, project_id, session_id
-      };
+      const shared = await rehostLocalPaths(client, {
+        model, source_images, reference_images, aspect_ratio, enhance_prompt, font_ids,
+        visual_dna_ids, moodboard_id, enable_web_search, resolution, quality, background, output_format, output_compression, moderation, mask_image_url, preset_id, cinematic, skip_color_palette, project_id, session_id
+      }, { allowLocalFiles: !options.remote });
+      source_images = shared.source_images;
+      reference_images = [...(shared.reference_images || []), shared.mask_image_url].filter(Boolean);
       const settings = imageSettings(shared);
 
       // Batch mode: N different edit instructions against the SAME source
@@ -449,7 +453,7 @@ function registerGenerateTools(server, client, options = {}) {
           status_args: { generation_ids: batch.ids, wait: true },
           reference_images: [...(source_images || []), ...(reference_images || [])]
         });
-        return pollBatch(client, batch, { interval: (batch.ok[0].gen.poll_interval_hint || 3) * 1000, timeout: 150000 }, 'generate_image_edit', model);
+        return pollBatch(client, batch, { interval: (batch.ok[0].gen.poll_interval_hint || 3) * 1000, timeout: 150000 }, 'generate_image_edit', model, { settings, reference_images: [...(source_images || []), ...reference_images] });
       }
 
       const gen = await client.post('/v1/generate/image-edit', { ...shared, prompt, num_images });
@@ -1441,7 +1445,14 @@ function registerGenerateTools(server, client, options = {}) {
           // the live card merges this payload over its submit-time state, so
           // the finished card shows every reference, including server-side
           // additions (DNA stills, merged sources) the tool call never named.
-          reference_images: res.reference_images,
+          reference_images: res.reference_images,
+          reference_videos: res.reference_videos,
+          reference_audio: res.reference_audio,
+          ...(res.preset ? { settings: {
+            preset_id: res.preset.id,
+            preset_name: res.preset.name,
+            preset_thumbnail: res.preset.thumbnail_url,
+          } } : {}),
           // Transcription results ride the same status tool as media
           // generations; without these the transcript widget merges a payload
           // with no transcript in it.