Image Generator MCP Server

Instructions

Implementation Reference

• src/index.ts:74-104 (handler)

  Handler for CallToolRequestSchema specifically for the 'generate_image' tool. Validates the tool name and arguments, generates the image using ImageGenerator, saves the file using FileSaver, and returns the file URI as tool result.

  this.server.setRequestHandler(
    CallToolRequestSchema,
    async (request) => {
      if (request.params.name !== "generate_image") {
        throw new McpError(
          ErrorCode.MethodNotFound,
          `Unknown tool: ${request.params.name}`
        );
      }
  
      if (!isValidImageGenerationArgs(request.params.arguments)) {
        throw new McpError(
          ErrorCode.InvalidParams,
          "Invalid image generation arguments"
        )
      };
      
      const { prompt, imageName } = request.params.arguments;
      const base64 = await new ImageGenerator().generateImage(prompt);
      const fileName = `${imageName.replace(/\..*$/, '')}.png`;
      const filepath = await imageSaver.saveBase64(fileName, base64!);
  
      return {
        toolResult: {
          uri: `file://${filepath}`,
          type: 'image',
          mimeType: 'image/png'
        }
      }
    }
  )

• src/types.ts:1-13 (schema)

  Type definition for image generation parameters and validation function used in the handler to check input arguments.

  export interface ImageGenerationRequestParams {
      prompt: string;
      imageName: string;
  }
  
  export function isValidImageGenerationArgs(args: any): args is ImageGenerationRequestParams {
      return typeof args === "object" &&
          args !== null &&
          "prompt" in args &&
          typeof args.prompt === 'string' &&
          "imageName" in args &&
          typeof args.imageName === 'string';  
  }

• src/index.ts:53-71 (registration)

  Registers the 'generate_image' tool in the ListToolsRequestHandler with description and input schema.

    tools: [{
      name: "generate_image",
      description: "Generate an image from a prompt.",
      inputSchema: {
        type: "object",
        properties: {
          prompt: {
            type: "string",
            description: "A prompt detailing what image to generate."
          },
          imageName: {
            type: "string",
            description: "The filename for the image excluding any extensions."
          }
        },
        required: ["prompt", "imageName"]
      }
    }]
  })

• src/image-generator.ts:12-20 (helper)

  Core image generation logic using OpenAI's DALL-E 3 model to generate base64 image from prompt.

  async generateImage(prompt: string, size: ImageGenerateParams['size'] = "1024x1024") {
      const response = await this.openai.images.generate({
          model: IMAGE_MODEL,
          prompt,
          size,
          response_format: 'b64_json'
      });
      return response.data[0].b64_json;
  }