Merge pull request #3364 from Kilo-Org/beatlevic/cleanup-llm-benchmark

Cleanup and fix llm benchmark

Author: beatlevic

apps/kilocode-docs/docs/cli.md

@@ -73,9 +73,11 @@ async function askUserApproval(category: string, testName: string, input: string
 		console.log("─".repeat(80))
 		console.log("\n" + "─".repeat(80))
 
-		rl.question("\nIs this acceptable? (y/n): ", (answer) => {
+		rl.question("\nIs this acceptable? [Y/n]: ", (answer) => {
 			rl.close()
-			resolve(answer.toLowerCase() === "y" || answer.toLowerCase() === "yes")
+			const trimmed = answer.trim().toLowerCase()
+			const isApproved = trimmed === "" || trimmed === "y" || trimmed === "yes"
+			resolve(isApproved)
 		})
 	})
 }

src/test-llm-autocompletion/approvals.ts

@@ -0,0 +1,7 @@
+async function fetchWithRetry(url, maxRetries = 3) {
+	for (let attempt = 0; attempt < maxRetries; attempt++) {
+		try {
+			return await fetch(url);
+		} catch (error) {
+			if (attempt === maxRetries - 1) throw error;
+			await new Promise(resolve => setTimeout(resolve, 1000)]);