imagegen-mcp-server/dist/server.js

MCP server for AI image generation with OpenAI DALL-E, Google Imagen, Gemini, and Flux via Replicate

#!/usr/bin/env node
import "dotenv/config";
import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";
import { generateImageOpenAI } from "./providers/openai.js";
import { generateImageGoogle } from "./providers/google.js";
import { generateImageGemini } from "./providers/gemini.js";
import { generateImageReplicate } from "./providers/replicate.js";
import { z } from "zod";
const server = new McpServer({
    name: "imagegen-mcp-server",
    version: "0.1.0",
});
server.registerTool("image.generate.openai", {
    description: "Generate an image using OpenAI (default model gpt-image-1). Returns a saved file path and optional base64.",
    inputSchema: {
        prompt: z.string(),
        negativePrompt: z.string().optional(),
        width: z.number().int().positive().optional(),
        height: z.number().int().positive().optional(),
        size: z.string().optional(),
        format: z.enum(["png", "jpeg", "jpg", "webp"]).optional(),
        seed: z.number().int().optional(),
        quality: z.enum(["standard", "hd"]).optional(),
        style: z.string().optional(),
        background: z.enum(["transparent", "solid"]).optional(),
        model: z.string().optional(),
        returnBase64: z.boolean().optional(),
        filenameHint: z.string().optional(),
    },
}, async (args) => {
    const r = await generateImageOpenAI(args);
    const parts = [];
    parts.push({
        type: "text",
        text: `provider=openai model=${r.model ?? ""} saved=${r.path}`.trim(),
    });
    if (r.base64)
        parts.push({ type: "image", data: r.base64, mimeType: r.mimeType });
    return { content: parts };
});
server.registerTool("image.generate.google", {
    description: "Generate an image using Google (e.g., Imagen 3). Requires GOOGLE_API_KEY and GOOGLE_IMAGEN_ENDPOINT. Returns a saved file path and optional base64.",
    inputSchema: {
        prompt: z.string(),
        negativePrompt: z.string().optional(),
        width: z.number().int().positive().optional(),
        height: z.number().int().positive().optional(),
        size: z.string().optional(),
        format: z.enum(["png", "jpeg", "jpg", "webp"]).optional(),
        seed: z.number().int().optional(),
        quality: z.string().optional(),
        style: z.string().optional(),
        background: z.enum(["transparent", "solid"]).optional(),
        model: z.string().optional(),
        returnBase64: z.boolean().optional(),
        filenameHint: z.string().optional(),
    },
}, async (args) => {
    const r = await generateImageGoogle(args);
    const parts = [];
    parts.push({
        type: "text",
        text: `provider=google model=${r.model ?? ""} saved=${r.path}`.trim(),
    });
    if (r.base64)
        parts.push({ type: "image", data: r.base64, mimeType: r.mimeType });
    return { content: parts };
});
server.registerTool("image.generate.gemini", {
    description: "Generate an image using Google Gemini via @google/genai (default gemini-2.5-flash-image-preview). Requires GOOGLE_API_KEY.",
    inputSchema: {
        prompt: z.string(),
        model: z.string().optional(),
        returnBase64: z.boolean().optional(),
        filenameHint: z.string().optional(),
    },
}, async (args) => {
    const r = await generateImageGemini(args);
    const parts = [];
    parts.push({
        type: "text",
        text: `provider=google(gemini) model=${r.model ?? ""} saved=${r.path}`.trim(),
    });
    if (r.base64)
        parts.push({ type: "image", data: r.base64, mimeType: r.mimeType });
    return { content: parts };
});
server.registerTool("image.generate.replicate", {
    description: "Generate an image using Replicate models: Flux 1.1 Pro (default), Qwen Image, or SeedDream-4. Requires REPLICATE_API_TOKEN.",
    inputSchema: {
        prompt: z.string(),
        model: z.string().optional(),
        width: z.number().int().positive().optional(),
        height: z.number().int().positive().optional(),
        size: z.string().optional(),
        format: z.enum(["png", "jpeg", "jpg", "webp"]).optional(),
        seed: z.number().int().optional(),
        returnBase64: z.boolean().optional(),
        filenameHint: z.string().optional(),
    },
}, async (args) => {
    const r = await generateImageReplicate(args);
    const parts = [];
    parts.push({
        type: "text",
        text: `provider=replicate model=${r.model ?? ""} saved=${r.path}`.trim(),
    });
    if (r.base64)
        parts.push({ type: "image", data: r.base64, mimeType: r.mimeType });
    return { content: parts };
});
async function main() {
    const transport = new StdioServerTransport();
    await server.connect(transport);
}
main().catch((err) => {
    console.error("Fatal error starting server:", err);
    process.exit(1);
});