MCP Kling

Instructions

Implementation Reference

• src/kling-client.ts:334-374 (handler)

  Core handler function that validates input, processes image URLs, calls Kling AI /v1/videos/effects API, and returns task_id.
  async applyVideoEffect(request: VideoEffectsRequest): Promise<{ task_id: string }> { const path = '/v1/videos/effects'; // Validate image count based on effect type const dualCharacterEffects = ['hug', 'kiss', 'heart_gesture']; const singleImageEffects = ['squish', 'expansion', 'fuzzyfuzzy', 'bloombloom', 'dizzydizzy']; if (dualCharacterEffects.includes(request.effect_scene) && request.image_urls.length !== 2) { throw new Error(`Effect "${request.effect_scene}" requires exactly 2 images`); } if (singleImageEffects.includes(request.effect_scene) && request.image_urls.length !== 1) { throw new Error(`Effect "${request.effect_scene}" requires exactly 1 image`); } // Process all image URLs const processedImageUrls = await Promise.all( request.image_urls.map(url => this.processImageUrl(url)) ); const body: any = { input: { image_urls: processedImageUrls.filter(url => url !== undefined), effect_scene: request.effect_scene, duration: request.duration || '5', } }; // Always add model_name body.input.model_name = request.model_name || 'kling-v2-master'; try { const response = await this.axiosInstance.post(path, body); return response.data.data; } catch (error) { if (axios.isAxiosError(error)) { throw new Error(`Kling API error: ${error.response?.data?.message || error.message}`); } throw error; } }
• src/kling-client.ts:73-78 (schema)

  TypeScript interface defining the input parameters for the applyVideoEffect function.
  export interface VideoEffectsRequest { image_urls: string[]; effect_scene: 'hug' | 'kiss' | 'heart_gesture' | 'squish' | 'expansion' | 'fuzzyfuzzy' | 'bloombloom' | 'dizzydizzy'; duration?: '5' | '10'; model_name?: 'kling-v1' | 'kling-v1.5' | 'kling-v1.6' | 'kling-v2-master'; }
• src/index.ts:296-325 (registration)

  MCP tool registration object defining name, description, and input schema for apply_video_effect.
  name: 'apply_video_effect', description: 'Apply pre-defined animation effects to static images using Kling AI. Create emotionally expressive videos from portraits with effects like hugging, kissing, or playful animations. Dual-character effects (hug, kiss, heart_gesture) require exactly 2 images. Single-image effects (squish, expansion, fuzzyfuzzy, bloombloom, dizzydizzy) require 1 image. Perfect for social media content and creative storytelling.', inputSchema: { type: 'object', properties: { image_urls: { type: 'array', items: { type: 'string', }, description: 'Array of image URLs. Use 2 images for hug/kiss/heart_gesture effects, 1 image for squish/expansion/fuzzyfuzzy/bloombloom/dizzydizzy effects', }, effect_scene: { type: 'string', enum: ['hug', 'kiss', 'heart_gesture', 'squish', 'expansion', 'fuzzyfuzzy', 'bloombloom', 'dizzydizzy'], description: 'The animation effect to apply. Dual-character: hug, kiss, heart_gesture. Single-image: squish, expansion, fuzzyfuzzy, bloombloom, dizzydizzy', }, duration: { type: 'string', enum: ['5', '10'], description: 'Video duration in seconds (default: 5)', }, model_name: { type: 'string', enum: ['kling-v1', 'kling-v1.5', 'kling-v1.6', 'kling-v2-master'], description: 'Model version to use (default: kling-v2-master)', }, }, required: ['image_urls', 'effect_scene'], },
• src/index.ts:606-624 (handler)

  MCP server request handler case that parses arguments, calls KlingClient.applyVideoEffect, and formats response.
  case 'apply_video_effect': { const effectRequest: VideoEffectsRequest = { image_urls: args.image_urls as string[], effect_scene: args.effect_scene as 'hug' | 'kiss' | 'heart_gesture' | 'squish' | 'expansion' | 'fuzzyfuzzy' | 'bloombloom' | 'dizzydizzy', duration: (args.duration as '5' | '10') || '5', model_name: (args.model_name as 'kling-v1' | 'kling-v1.5' | 'kling-v1.6' | 'kling-v2-master' | undefined) || 'kling-v2-master', }; const result = await klingClient.applyVideoEffect(effectRequest); return { content: [ { type: 'text', text: `Video effect "${effectRequest.effect_scene}" applied successfully!\nTask ID: ${result.task_id}\n\nThe effect video is being generated.\nUse the check_video_status tool with this task ID to check the progress.`, }, ], }; }