generate_image

Instructions

Implementation Reference

• src/tools/generate.ts:84-101 (handler)

  The async handler function that executes the 'generate_image' tool logic. Calls client.generate() with prompt, negative_prompt, width, height, steps, cfg, seed, and checkpoint, then returns image URLs.

    async (args) => {
      const result = await client.generate({
        prompt: args.prompt,
        negativePrompt: args.negative_prompt,
        width: args.width,
        height: args.height,
        steps: args.steps,
        cfg: args.cfg,
        seed: args.seed,
        checkpoint: args.checkpoint,
      });
  
      return textResult(
        `Generated ${result.images.length} image(s) (prompt_id: ${result.promptId}):`,
        result.images,
      );
    },
  );

• src/tools/generate.ts:6-44 (schema)

  Zod schema defining input parameters for the 'generate_image' tool: prompt (string), negative_prompt (optional string), width (64-2048, default 1024), height (64-2048, default 1024), steps (1-150, default 25), cfg (1-30, default 7), seed (optional int), checkpoint (optional string).

  const generateImageSchema = {
    prompt: z
      .string()
      .min(1)
      .describe("Text prompt describing the image to generate"),
    negative_prompt: z.string().optional().describe("What to avoid in the image"),
    width: z
      .number()
      .int()
      .min(64)
      .max(2048)
      .default(1024)
      .describe("Image width in pixels"),
    height: z
      .number()
      .int()
      .min(64)
      .max(2048)
      .default(1024)
      .describe("Image height in pixels"),
    steps: z
      .number()
      .int()
      .min(1)
      .max(150)
      .default(25)
      .describe("Number of diffusion steps"),
    cfg: z
      .number()
      .min(1)
      .max(30)
      .default(7)
      .describe("CFG / prompt adherence (1-30)"),
    seed: z.number().int().optional().describe("Seed for reproducibility"),
    checkpoint: z
      .string()
      .optional()
      .describe("Checkpoint filename (defaults to COMFYUI_DEFAULT_CKPT)"),
  };

• src/tools/generate.ts:76-83 (registration)

  Registration of the 'generate_image' tool via server.tool() with name 'generate_image' and description. Part of the registerGenerateTools() function called from src/server.ts:42.

  export function registerGenerateTools(
    server: McpServer,
    client: ComfyUIClient,
  ): void {
    server.tool(
      "generate_image",
      "Generate an image from a text prompt using ComfyUI's default txt2img workflow. Returns one or more image URLs served directly by the ComfyUI instance.",
      generateImageSchema,

• src/server.ts:42-52 (registration)

  Call site where registerGenerateTools() is invoked with the MCP server and ComfyUIClient instance to register the 'generate_image' tool.

      registerGenerateTools(s, client);
      registerRefineTool(s, client);
      registerUpscaleTool(s, client);
      registerModelTools(s, client);
      registerImageTools(s, client);
      registerConditioningTools(s, client);
      registerTemplateTools(s, client, templateStore);
      return s;
    };
    return { client, buildServer };
  }

• src/comfyui/client.ts:47-59 (helper)

  The generate() method on ComfyUIClient that the handler calls. It builds a txt2img workflow via the txt2img() helper and runs it via runWorkflow().

  async generate(params: GenerateParams): Promise<GenerateResult> {
    const workflow = txt2img({
      prompt: params.prompt,
      negativePrompt: params.negativePrompt ?? "",
      width: params.width ?? 1024,
      height: params.height ?? 1024,
      steps: params.steps ?? 25,
      cfg: params.cfg ?? 7,
      seed: params.seed ?? Math.floor(Math.random() * 2 ** 32),
      checkpoint: params.checkpoint ?? DEFAULT_CHECKPOINT,
    });
    return this.runWorkflow(workflow);
  }

• src/comfyui/workflows.ts:30-72 (helper)

  The txt2img() workflow builder that creates the ComfyUI node graph (KSampler, CheckpointLoaderSimple, EmptyLatentImage, CLIPTextEncode, VAEDecode, SaveImage) used by the 'generate_image' tool.

  export function txt2img(params: Txt2ImgParams): Workflow {
    return {
      "3": {
        class_type: "KSampler",
        inputs: {
          seed: params.seed,
          steps: params.steps,
          cfg: params.cfg,
          sampler_name: "euler",
          scheduler: "normal",
          denoise: 1,
          model: ["4", 0],
          positive: ["6", 0],
          negative: ["7", 0],
          latent_image: ["5", 0],
        },
      },
      "4": {
        class_type: "CheckpointLoaderSimple",
        inputs: { ckpt_name: params.checkpoint },
      },
      "5": {
        class_type: "EmptyLatentImage",
        inputs: { width: params.width, height: params.height, batch_size: 1 },
      },
      "6": {
        class_type: "CLIPTextEncode",
        inputs: { text: params.prompt, clip: ["4", 1] },
      },
      "7": {
        class_type: "CLIPTextEncode",
        inputs: { text: params.negativePrompt, clip: ["4", 1] },
      },
      "8": {
        class_type: "VAEDecode",
        inputs: { samples: ["3", 0], vae: ["4", 2] },
      },
      "9": {
        class_type: "SaveImage",
        inputs: { filename_prefix: "comfyui-mcp", images: ["8", 0] },
      },
    };
  }