FAL Image/Video MCP Server

Instructions

Implementation Reference

• src/index.ts:495-563 (handler)

  Core handler function for flux_kontext (image generation model). Prepares parameters (including flux-specific steps/guidance), calls fal.subscribe on the specific endpoint, processes images with download/dataUrl/autoopen, and returns formatted result.

  private async handleImageGeneration(args: any, model: any) {
    const {
      prompt,
      image_size = 'landscape_4_3',
      num_inference_steps = 25,
      guidance_scale = 3.5,
      num_images = 1,
      negative_prompt,
      safety_tolerance,
      raw,
    } = args;
  
    try {
      // Configure FAL client lazily with query config override
      configureFalClient(this.currentQueryConfig);
      const inputParams: any = { prompt };
      
      // Add common parameters
      if (image_size) inputParams.image_size = image_size;
      if (num_images > 1) inputParams.num_images = num_images;
      
      // Add model-specific parameters based on model capabilities
      if (model.id.includes('flux') || model.id.includes('stable_diffusion')) {
        if (num_inference_steps) inputParams.num_inference_steps = num_inference_steps;
        if (guidance_scale) inputParams.guidance_scale = guidance_scale;
      }
      if ((model.id.includes('stable_diffusion') || model.id === 'ideogram_v3') && negative_prompt) {
        inputParams.negative_prompt = negative_prompt;
      }
      if (model.id.includes('flux_pro') && safety_tolerance) {
        inputParams.safety_tolerance = safety_tolerance;
      }
      if (model.id === 'flux_pro_ultra' && raw !== undefined) {
        inputParams.raw = raw;
      }
  
      const result = await fal.subscribe(model.endpoint, { input: inputParams });
      const imageData = result.data as FalImageResult;
  
      const processedImages = await downloadAndProcessImages(imageData.images, model.id);
  
      return {
        content: [
          {
            type: 'text',
            text: JSON.stringify({
              model: model.name,
              id: model.id,
              endpoint: model.endpoint,
              prompt,
              images: processedImages,
              metadata: inputParams,
              download_path: DOWNLOAD_PATH,
              data_url_settings: {
                enabled: ENABLE_DATA_URLS,
                max_size_mb: Math.round(MAX_DATA_URL_SIZE / 1024 / 1024),
              },
              autoopen_settings: {
                enabled: AUTOOPEN,
                note: AUTOOPEN ? "Files automatically opened with default application" : "Auto-open disabled"
              },
            }, null, 2),
          },
        ],
      };
    } catch (error) {
      throw new Error(`${model.name} generation failed: ${error}`);
    }
  }

• src/index.ts:102-102 (registration)

  Registration of the flux_kontext tool in MODEL_REGISTRY.imageGeneration array, defining its ID, FAL endpoint, name, and description.

  { id: 'flux_kontext', endpoint: 'fal-ai/flux-pro/kontext/text-to-image', name: 'FLUX Kontext Pro', description: 'State-of-the-art prompt adherence and typography' },

• src/index.ts:346-393 (schema)

  Dynamic input schema generation for flux_kontext tool (name=model.id). For imageGeneration category: defines required 'prompt', optional image_size, num_images, flux-specific num_inference_steps/guidance_scale.

  private generateToolSchema(model: any, category: string) {
    const baseSchema = {
      name: model.id,
      description: `${model.name} - ${model.description}`,
      inputSchema: {
        type: 'object',
        properties: {} as any,
        required: [] as string[],
      },
    };
  
    if (category === 'imageGeneration') {
      baseSchema.inputSchema.properties = {
        prompt: { type: 'string', description: 'Text prompt for image generation' },
        image_size: { type: 'string', enum: ['square_hd', 'square', 'portrait_4_3', 'portrait_16_9', 'landscape_4_3', 'landscape_16_9'], default: 'landscape_4_3' },
        num_images: { type: 'number', default: 1, minimum: 1, maximum: 4 },
      };
      baseSchema.inputSchema.required = ['prompt'];
      
      // Add model-specific parameters
      if (model.id.includes('flux') || model.id.includes('stable_diffusion')) {
        baseSchema.inputSchema.properties.num_inference_steps = { type: 'number', default: 25, minimum: 1, maximum: 50 };
        baseSchema.inputSchema.properties.guidance_scale = { type: 'number', default: 3.5, minimum: 1, maximum: 20 };
      }
      if (model.id.includes('stable_diffusion') || model.id === 'ideogram_v3') {
        baseSchema.inputSchema.properties.negative_prompt = { type: 'string', description: 'Negative prompt' };
      }
    } else if (category === 'textToVideo') {
      baseSchema.inputSchema.properties = {
        prompt: { type: 'string', description: 'Text prompt for video generation' },
        duration: { type: 'number', default: 5, minimum: 1, maximum: 30 },
        aspect_ratio: { type: 'string', enum: ['16:9', '9:16', '1:1', '4:3', '3:4'], default: '16:9' },
      };
      baseSchema.inputSchema.required = ['prompt'];
    } else if (category === 'imageToVideo') {
      baseSchema.inputSchema.properties = {
        image_url: { type: 'string', description: 'URL of the input image' },
        prompt: { type: 'string', description: 'Motion description prompt' },
        duration: { type: 'string', enum: ['5', '10'], default: '5', description: 'Video duration in seconds' },
        aspect_ratio: { type: 'string', enum: ['16:9', '9:16', '1:1'], default: '16:9' },
        negative_prompt: { type: 'string', description: 'What to avoid in the video' },
        cfg_scale: { type: 'number', default: 0.5, minimum: 0, maximum: 1, description: 'How closely to follow the prompt' }
      };
      baseSchema.inputSchema.required = ['image_url', 'prompt'];
    }
  
    return baseSchema;
  }

• src/index.ts:399-402 (registration)

  Registration loop in listTools handler that adds flux_kontext schema to the tools list using generateToolSchema.

  // Generate tools for each category
  for (const model of MODEL_REGISTRY.imageGeneration) {
    tools.push(this.generateToolSchema(model, 'imageGeneration'));
  }

• src/index.ts:476-478 (handler)

  Dispatch logic in CallToolRequestSchema handler that routes flux_kontext calls (matching imageGeneration) to the handleImageGeneration function.

  if (MODEL_REGISTRY.imageGeneration.find(m => m.id === name)) {
    return await this.handleImageGeneration(args, model);
  } else if (MODEL_REGISTRY.textToVideo.find(m => m.id === name)) {