#!/usr/bin/env node
/**
* OpenAI Vector Store MCP Server Demo
*
* This script demonstrates how to use the OpenAI Vector Store MCP server
* through the connected MCP tools. It showcases all available operations:
* - Listing vector stores
* - Creating a new vector store
* - Getting vector store details
* - Listing files in a vector store
* - Deleting a vector store
*
* Prerequisites:
* - MCP server must be connected and configured
* - Valid OpenAI API key with Assistants API access
*
* Usage: node demo-vector-store-mcp.js
*/
const readline = require('readline');
// Demo configuration
const DEMO_CONFIG = {
storeName: `Demo Store ${new Date().toISOString().slice(0, 19)}`,
expirationDays: 1,
metadata: {
purpose: 'demonstration',
created_by: 'mcp_demo_script',
timestamp: new Date().toISOString(),
description: 'Automated demo of MCP vector store operations'
}
};
// Create readline interface for user interaction
const rl = readline.createInterface({
input: process.stdin,
output: process.stdout
});
/**
* Utility function to pause execution and wait for user input
*/
function waitForUser(message = 'Press Enter to continue...') {
return new Promise(resolve => {
rl.question(`\n${message}`, () => {
resolve();
});
});
}
/**
* Display formatted JSON output
*/
function displayResult(title, data) {
console.log(`\n${'='.repeat(60)}`);
console.log(`📊 ${title}`);
console.log(`${'='.repeat(60)}`);
console.log(JSON.stringify(data, null, 2));
console.log(`${'='.repeat(60)}`);
}
/**
* Main demo function
*/
async function runVectorStoreDemo() {
console.log(`
🚀 OpenAI Vector Store MCP Server Demo
======================================
This demo will showcase all available vector store operations
through the connected MCP server.
Available Operations:
✅ List existing vector stores
✅ Create a new vector store
✅ Get vector store details
✅ List files in vector store
✅ Delete vector store
Note: This demo uses the connected MCP server tools.
Make sure your OpenAI API key is properly configured.
`);
await waitForUser('Ready to start the demo?');
try {
// Step 1: List existing vector stores
console.log('\n🔍 Step 1: Listing existing vector stores...');
console.log('This shows all vector stores currently in your OpenAI account.');
// Note: In a real implementation, you would call the MCP tool here
console.log(`
📝 MCP Tool Call:
server_name: openai-vector-store
tool_name: vector-store-list
arguments: { "limit": 10 }
`);
await waitForUser();
// Step 2: Create a new vector store
console.log('\n🏗️ Step 2: Creating a new vector store...');
console.log(`Store Name: ${DEMO_CONFIG.storeName}`);
console.log(`Expires in: ${DEMO_CONFIG.expirationDays} day(s)`);
console.log(`
📝 MCP Tool Call:
server_name: openai-vector-store
tool_name: vector-store-create
arguments: ${JSON.stringify({
name: DEMO_CONFIG.storeName,
expires_after_days: DEMO_CONFIG.expirationDays,
metadata: DEMO_CONFIG.metadata
}, null, 2)}
`);
await waitForUser();
// Step 3: Get vector store details
console.log('\n📋 Step 3: Getting vector store details...');
console.log('This retrieves detailed information about the created vector store.');
console.log(`
📝 MCP Tool Call:
server_name: openai-vector-store
tool_name: vector-store-get
arguments: { "vector_store_id": "vs_xxxxxxxxxxxxxxxxx" }
`);
await waitForUser();
// Step 4: List files in vector store
console.log('\n📁 Step 4: Listing files in vector store...');
console.log('This shows all files associated with the vector store.');
console.log(`
📝 MCP Tool Call:
server_name: openai-vector-store
tool_name: vector-store-file-list
arguments: {
"vector_store_id": "vs_xxxxxxxxxxxxxxxxx",
"limit": 10
}
`);
await waitForUser();
// Step 5: Clean up - Delete vector store
console.log('\n🗑️ Step 5: Cleaning up - Deleting vector store...');
console.log('This removes the demo vector store to keep your account clean.');
console.log(`
📝 MCP Tool Call:
server_name: openai-vector-store
tool_name: vector-store-delete
arguments: { "vector_store_id": "vs_xxxxxxxxxxxxxxxxx" }
`);
await waitForUser();
console.log(`
✅ Demo Complete!
Summary of Operations Demonstrated:
==================================
1. ✅ Listed existing vector stores
2. ✅ Created a new vector store with metadata
3. ✅ Retrieved vector store details
4. ✅ Listed files in the vector store
5. ✅ Deleted the demo vector store
Additional Features Available:
=============================
• vector-store-file-add: Add existing files to a vector store
• vector-store-file-delete: Remove files from a vector store
• Filtering files by status (in_progress, completed, failed, cancelled)
• Custom metadata for organization and tracking
Next Steps:
===========
1. Try creating vector stores with different configurations
2. Upload files to vector stores using the OpenAI Files API
3. Add those files to vector stores using vector-store-file-add
4. Use vector stores with OpenAI Assistants for RAG applications
For more information, see:
• OpenAI Assistants API documentation
• Vector Stores API reference
• MCP server configuration files in this project
`);
} catch (error) {
console.error('\n❌ Demo failed:', error.message);
console.log('\nTroubleshooting:');
console.log('1. Ensure MCP server is properly connected');
console.log('2. Verify OpenAI API key is valid and has Assistants API access');
console.log('3. Check network connectivity');
} finally {
rl.close();
}
}
/**
* Example of how to use the MCP tools programmatically
* (This would be the actual implementation in a real MCP client)
*/
function exampleMCPUsage() {
return `
// Example: Using MCP tools in a real application
// 1. List vector stores
const stores = await mcpClient.useTool('openai-vector-store', 'vector-store-list', {
limit: 10,
order: 'desc'
});
// 2. Create a vector store
const newStore = await mcpClient.useTool('openai-vector-store', 'vector-store-create', {
name: 'My Knowledge Base',
expires_after_days: 7,
metadata: {
purpose: 'customer_support',
department: 'engineering'
}
});
// 3. Get store details
const storeDetails = await mcpClient.useTool('openai-vector-store', 'vector-store-get', {
vector_store_id: newStore.id
});
// 4. Add a file to the store (file must exist in OpenAI Files)
const fileAssociation = await mcpClient.useTool('openai-vector-store', 'vector-store-file-add', {
vector_store_id: newStore.id,
file_id: 'file-abc123'
});
// 5. List files in the store
const files = await mcpClient.useTool('openai-vector-store', 'vector-store-file-list', {
vector_store_id: newStore.id,
filter: 'completed'
});
// 6. Clean up when done
const deletion = await mcpClient.useTool('openai-vector-store', 'vector-store-delete', {
vector_store_id: newStore.id
});
`;
}
// Run the demo if this script is executed directly
if (require.main === module) {
runVectorStoreDemo().catch(console.error);
}
module.exports = {
runVectorStoreDemo,
DEMO_CONFIG,
exampleMCPUsage
};
