#!/usr/bin/env node
import { Client } from '@modelcontextprotocol/sdk/client/index.js';
import { StdioClientTransport } from '@modelcontextprotocol/sdk/client/stdio.js';
// Test cases for natural language queries
const testQueries = [
// IP Core queries
"Find DDR5 PHY IP vendors for 7nm process",
"Show me USB 3.0 IP cores",
"I need PCIe 5.0 controller IP",
"List memory interface IPs for 5nm",
// ASIC Service queries
"Find ASIC design services",
"Show verification services for 7nm designs",
"I need backend design services",
// Price estimation queries
"Estimate ASIC NRE cost for 7nm",
"What's the mask cost for 5nm?",
"Calculate IP licensing cost for DDR5 PHY",
// Comparison queries
"Compare TSMC vs Samsung foundry services",
"Which is better: Synopsys or Cadence for IP?",
// Complex queries
"Find DDR5 IP vendors that support 7nm TSMC process with low power options",
"I'm designing an AI accelerator at 7nm, what services do I need?",
// Ambiguous query to test suggestions
"I need something for my chip"
];
async function testNaturalLanguageQueries() {
console.log('š Starting MCP Server Test for Natural Language Queries\n');
// Create transport connected to the server
const serverPath = new URL('../dist/index.js', import.meta.url).pathname;
const transport = new StdioClientTransport({
command: 'node',
args: [serverPath]
});
const client = new Client({
name: 'test-nlp-client',
version: '1.0.0'
}, {
capabilities: {}
});
try {
await client.connect(transport);
console.log('ā
Connected to MCP server\n');
// Test natural language queries
console.log('š Testing Natural Language Queries:\n');
for (const query of testQueries) {
console.log(`\nš Query: "${query}"`);
console.log('-'.repeat(60));
try {
const result = await client.callTool('natural_language_query', {
query: query
});
const response = JSON.parse(result.content[0].text);
if (response.query_info) {
console.log(`ā
Intent: ${response.query_info.understood_intent}`);
console.log(`š Confidence: ${(response.query_info.confidence * 100).toFixed(1)}%`);
console.log(`š§ Parameters: ${JSON.stringify(response.query_info.extracted_parameters)}`);
if (response.query_info.suggestions_for_improvement) {
console.log(`š” Suggestions:`);
response.query_info.suggestions_for_improvement.forEach((s: string) =>
console.log(` ${s}`)
);
}
}
if (response.status === 'need_clarification') {
console.log(`ā ļø Need clarification: ${response.message}`);
} else if (response.result) {
console.log(`š Results: Found ${response.result.vendors?.length || response.result.services?.length || 0} items`);
}
} catch (error) {
console.error(`ā Error: ${error}`);
}
}
// Test resources
console.log('\n\nš Testing Resources:\n');
const resources = await client.listResources();
console.log(`Found ${resources.resources.length} resources:`);
resources.resources.forEach((r: any) => {
console.log(` - ${r.name}: ${r.uri}`);
});
// Read glossary
console.log('\nš Reading Glossary Sample:');
const glossary = await client.readResource('semiconductor://glossary');
const glossaryData = JSON.parse(glossary.contents[0].text);
console.log(`Total terms: ${glossaryData.total_terms}`);
console.log(`Categories: ${Object.keys(glossaryData.categories).join(', ')}`);
// Read process nodes
console.log('\nš Reading Process Nodes:');
const processNodes = await client.readResource('semiconductor://process-nodes');
const nodesData = JSON.parse(processNodes.contents[0].text);
console.log(`Available nodes: ${nodesData.available_nodes.join(', ')}`);
} catch (error) {
console.error('ā Test failed:', error);
} finally {
await client.close();
console.log('\nā
Test completed');
process.exit(0);
}
}
// Run the test
testNaturalLanguageQueries().catch(console.error);
