fix(eval): use percentages in evaluation summary

apps/cli/src/commands/eval/statistics.ts

@@ -181,7 +181,7 @@ export function calculateEvaluationSummary(
 }
 
 function formatScore(value: number): string {
-  return value.toFixed(3);
+  return `${Math.round(value * 100)}%`;
 }
 
 export function formatEvaluationSummary(
@@ -226,7 +226,7 @@ export function formatEvaluationSummary(
   } else {
     overallVerdict = overallPassed ? 'PASS' : 'FAIL';
     verdictColor = overallPassed ? '\x1b[32m' : '\x1b[31m';
-    verdictText = `RESULT: ${overallVerdict}  (${summary.passedCount}/${summary.total} scored >= ${threshold}, mean: ${formatScore(summary.mean)})`;
+    verdictText = `RESULT: ${overallVerdict}  (${summary.passedCount}/${summary.total} scored >= ${Math.round(threshold * 100)}%, mean: ${formatScore(summary.mean)})`;
   }
 
   lines.push('\n==================================================');
@@ -267,7 +267,7 @@ export function formatEvaluationSummary(
   lines.push('\nScore distribution:');
   for (const bin of summary.histogram) {
     const [start, end] = bin.range;
-    lines.push(`  ${start.toFixed(1)}-${end.toFixed(1)}: ${bin.count}`);
+    lines.push(`  ${Math.round(start * 100)}%-${Math.round(end * 100)}%: ${bin.count}`);
   }
 
   lines.push('\nTop performing tests:');

apps/cli/test/eval.integration.test.ts

@@ -208,7 +208,7 @@ describe('agentv eval CLI', () => {
 
       // Don't check stderr - it may contain stack traces or other diagnostics
       expect(stdout).toContain('Using target (test-file): file-target [provider=mock]');
-      expect(stdout).toContain('Mean score: 0.750');
+      expect(stdout).toContain('Mean score: 75%');
       // Std deviation is an implementation detail - don't check it
 
       const outputPath = extractOutputPath(stdout);

apps/cli/test/unit/matrix-summary.test.ts

@@ -34,8 +34,8 @@ describe('formatMatrixSummary', () => {
     expect(output).toContain('test-1');
     expect(output).toContain('test-2');
     expect(output).toContain('Average');
-    expect(output).toContain('0.900');
-    expect(output).toContain('0.700');
+    expect(output).toContain('90%');
+    expect(output).toContain('70%');
   });
 
   it('handles missing test-target pairs with dash', () => {