HOW-TO-RUN.md

How to Run llmverify in Windsurf IDE

Step 1: Open Two Terminals

In Windsurf, press Ctrl+`` (backtick) to open the terminal. Click the +` button to open a second terminal tab.

Step 2: Terminal 1 - Start Server

cd "C:\Users\Subodh Kc\Desktop\App Building\Github\llmverify-npm"
npx llmverify-serve

Keep this running. You'll see:

Server started on http://localhost:9009

Step 3: Terminal 2 - Test It

cd "C:\Users\Subodh Kc\Desktop\App Building\Github\llmverify-npm"

# Test verification
$body = '{"content": "This is a test AI response"}' 
Invoke-RestMethod -Uri "http://localhost:9009/verify" -Method POST -ContentType "application/json" -Body $body

You'll get a JSON response with risk scores.

Step 4: Use in Your Code

const { createIDEExtension } = require('llmverify');

const verifier = createIDEExtension();
const result = await verifier.verify("AI response here");

console.log(result.verdict);    // [PASS] SAFE TO USE
console.log(result.riskScore);  // 6.3
console.log(result.safe);       // true

Quick Test

Run the example:

node examples/basic-usage.js

This will show you all the features working.

Common Issues

"Server not running"

Make sure Terminal 1 is still running with npx llmverify-serve

"Port in use"

npx llmverify-serve --port=8080

"Module not found"

npm install
npm run build

That's It!

Now you can verify AI responses in real-time while coding in Windsurf.