index.ts•4.4 kB
#!/usr/bin/env node
import { Server } from "@modelcontextprotocol/sdk/server/index.js";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";
import {
ListToolsRequestSchema,
CallToolRequestSchema,
ErrorCode,
McpError
} from "@modelcontextprotocol/sdk/types.js";
import OpenAI from "openai";
import dotenv from "dotenv";
dotenv.config();
const QWEN_BASE_URL = "https://dashscope-intl.aliyuncs.com/compatible-mode/v1";
const API_KEY = process.env.DASHSCOPE_API_KEY;
if (!API_KEY) {
throw new Error("DASHSCOPE_API_KEY environment variable is required");
}
interface QwenMaxArgs {
prompt: string;
max_tokens?: number;
temperature?: number;
}
class QwenMaxServer {
private server: Server;
private openai: OpenAI;
constructor() {
this.server = new Server(
{ name: "qwen_max", version: "1.0.0" },
{ capabilities: { tools: {} } }
);
this.openai = new OpenAI({
apiKey: API_KEY,
baseURL: QWEN_BASE_URL
});
this.setupHandlers();
this.setupErrorHandling();
}
private setupErrorHandling(): void {
this.server.onerror = (error: Error): void => {
console.error("[MCP Error]", error);
};
process.on("SIGINT", async (): Promise<void> => {
await this.server.close();
process.exit(0);
});
}
private setupHandlers(): void {
this.server.setRequestHandler(
ListToolsRequestSchema,
async () => ({
tools: [{
name: "qwen_max",
description: "Generate text using Qwen Max model",
inputSchema: {
type: "object",
properties: {
prompt: {
type: "string",
description: "The text prompt to generate content from"
},
max_tokens: {
type: "number",
description: "Maximum number of tokens to generate",
default: 8192
},
temperature: {
type: "number",
description: "Sampling temperature (0-2)",
default: 0.7,
minimum: 0,
maximum: 2
}
},
required: ["prompt"]
}
}]
})
);
this.server.setRequestHandler(
CallToolRequestSchema,
async (request) => {
if (request.params.name !== "qwen_max") {
throw new McpError(
ErrorCode.MethodNotFound,
`Unknown tool: ${request.params.name}`
);
}
const { prompt, max_tokens = 8192, temperature = 0.7 } =
request.params.arguments as QwenMaxArgs;
try {
const completion = await this.openai.chat.completions.create({
model: "qwen-max-latest",
messages: [{ role: "user", content: prompt }],
max_tokens,
temperature
});
return {
content: [{
type: "text",
text: completion.choices[0].message.content || ""
}]
};
} catch (error: any) {
console.error("Qwen API Error:", error);
throw new McpError(
ErrorCode.InternalError,
`Qwen API error: ${error.message}`
);
}
}
);
}
async run(): Promise<void> {
const transport = new StdioServerTransport();
await this.server.connect(transport);
console.error("Qwen Max MCP server running on stdio");
}
}
const server = new QwenMaxServer();
server.run().catch(console.error);