openaiclient.js•1.85 kB
require('dotenv').config();
const { OpenAI } = require("openai");
const axios = require("axios");
const openai = new OpenAI({ apiKey: process.env.OPENAI_API_KEY });
const mcpFunction = {
name: "get_ticket_context",
description: "Returns customer or org context from Zendesk using a ticket ID, user ID, or organization ID",
parameters: {
type: "object",
properties: {
ticket_id: { type: "integer", description: "Zendesk ticket ID" },
user_id: { type: "integer", description: "Zendesk user ID" },
organization_id: { type: "integer", description: "Zendesk org ID" }
},
required: []
}
};
(async () => {
const userMessage = "Can you give me context for ticket 6?";
const initialResponse = await openai.chat.completions.create({
model: "gpt-4-0613",
messages: [
{ role: "user", content: userMessage }
],
functions: [mcpFunction],
function_call: "auto"
});
const message = initialResponse.choices[0].message;
if (message.function_call) {
const { name, arguments: rawArgs } = message.function_call;
const args = JSON.parse(rawArgs);
// Call your MCP server
const mcpRes = await axios.post("http://localhost:3000/context", args);
const mcpData = mcpRes.data;
// Feed the result back to GPT to generate a final natural reply
const finalResponse = await openai.chat.completions.create({
model: "gpt-4-0613",
messages: [
{ role: "user", content: userMessage },
message,
{
role: "function",
name,
content: JSON.stringify(mcpData)
}
]
});
console.error("\n🔁 GPT-4 Final Response:\n");
console.error(finalResponse.choices[0].message.content);
} else {
// No function call needed
console.error("\nGPT-4 Response:\n", message.content);
}
})();
