@kolbo/mcp 1.84.3 → 1.85.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@kolbo/mcp",
-  "version": "1.84.3",
+  "version": "1.85.0",
   "description": "Kolbo AI MCP Server - Generate images, videos, music, speech, and sound effects from Claude Code",
   "main": "src/index.js",
   "bin": {

package/src/index.js

@@ -77,6 +77,7 @@ const { registerVoiceTools } = require('./tools/voices');
 const { registerMusicLibraryTools } = require('./tools/music_library');
 const { registerStockLibraryTools } = require('./tools/stock_library');
 const { registerAudioStemTools } = require('./tools/audio_stems');
+const { registerAnalyzeTools } = require('./tools/analyze');
 const { registerBlenderTools } = require('./tools/blender');
 const { registerApps, attachToolWidgetMeta } = require('./apps');
 const { attachToolAnnotations } = require('./toolAnnotations');
@@ -170,6 +171,7 @@ function createServer(opts = {}) {
   registerVisualDnaTools(server, client, toolOptions);
   registerMoodboardTools(server, client, toolOptions);
   registerColorPaletteTools(server, client, toolOptions);
+  registerAnalyzeTools(server, client, toolOptions);
   registerMediaTools(server, client, toolOptions);
   registerPresetTools(server, client, toolOptions);
   registerArtifactTools(server, client, toolOptions);

package/src/toolAnnotations.js

@@ -74,6 +74,7 @@ const DESTRUCTIVE_WRITE = [
   'chat_send_message', 'generate_character_sheet',
   'acquire_clean_music_track', 'import_music_track_to_library',
   'separate_audio_stems', 'clean_dialogue_leftovers', 'separate_ambience',
+  'analyze_video',
 
   // Deletes and whole-value replacement updates are conservatively destructive.
   'delete_voice', 'update_visual_dna', 'delete_visual_dna', 'delete_visual_dna_folder',

package/src/tools/_shared.js

@@ -55,7 +55,8 @@ const FILE_INPUT_TOOLS = [
   'generate_3d', 'edit_image', 'edit_video', 'transcribe_audio',
   'create_visual_dna', 'update_visual_dna', 'generate_character_sheet', 'clone_voice',
   'chat_send_message', 'create_moodboard', 'update_moodboard',
-  'separate_audio_stems', 'clean_dialogue_leftovers', 'separate_ambience'
+  'separate_audio_stems', 'clean_dialogue_leftovers', 'separate_ambience',
+  'analyze_video'
 ];
 
 const REMOTE_FILE_HINT =

package/src/tools/analyze.js

@@ -0,0 +1,50 @@
+/* ⛔ BACKWARD COMPATIBILITY: Tool names and arg names below are a PUBLIC
+ * CONTRACT. Never rename, remove, or break an existing tool/arg — old cached
+ * `npx @kolbo/mcp` installs in the wild will break silently. Add new tools or
+ * new OPTIONAL args only. Full rules: ../index.js top-of-file and CLAUDE.md. */
+
+const { z } = require('zod');
+const { projectIdField } = require('./_shared');
+
+function registerAnalyzeTools(server, client) {
+  // ─── analyze_video ──────────────────────────────────────────
+  // Kolbo's official video understanding. Sync — the server holds the socket for
+  // long videos, so there is nothing to poll.
+  server.tool(
+    'analyze_video',
+    'Understand a video with Kolbo\'s official video-understanding model (agentic Gemini: it navigates the timeline itself — frames, audio and transcript on demand — instead of sampling fixed frames, so long videos are cheap and timestamp / "when does X happen" / counting questions are answered directly). Pass a public https video URL or a YouTube URL plus an optional question in `prompt`; with no prompt you get an exhaustive description + verbatim transcript. Billed by the analyzer\'s real token usage. Local file → upload it first (create_upload_ticket / upload_media / media_upload_widget) and pass the returned URL. For subtitle files use transcribe_audio instead.',
+    {
+      video_url: z.string().optional().describe('Public https URL of the video (Kolbo media URL from upload_media / list_media, or any public file). Either this or youtube_url.'),
+      youtube_url: z.string().optional().describe('A youtube.com / youtu.be link. Either this or video_url.'),
+      prompt: z.string().optional().describe('The question or task about the video (e.g. "At what timestamp does the logo appear?", "Count how many people speak", "Summarize the three main arguments"). Omit for a full description + verbatim transcript.'),
+      quality: z.enum(['standard', 'hq']).optional().describe('"standard" (default, Flash-Lite) or "hq" (Flash, higher accuracy, ~2.5x the token price). Use hq for short clips where precision matters.'),
+      project_id: projectIdField,
+    },
+    async ({ video_url, youtube_url, prompt, quality, project_id }) => {
+      if (!video_url && !youtube_url) {
+        return { content: [{ type: 'text', text: 'Provide video_url or youtube_url. Local file? Upload it first (create_upload_ticket / upload_media / media_upload_widget) and pass the returned URL.' }], isError: true };
+      }
+      const result = await client.post('/v1/analyze/video', {
+        ...(video_url ? { video_url } : {}),
+        ...(youtube_url ? { youtube_url } : {}),
+        ...(prompt ? { prompt } : {}),
+        ...(quality ? { quality } : {}),
+        ...(project_id ? { project_id } : {}),
+      });
+      return {
+        content: [{
+          type: 'text',
+          text: JSON.stringify({
+            analysis: result.analysis,
+            model: result.model,
+            agentic: result.agentic,
+            usage: result.usage,
+            credits_used: result.credits_used,
+          }, null, 2),
+        }],
+      };
+    }
+  );
+}
+
+module.exports = { registerAnalyzeTools };