diff --git a/apps/cli/src/commands/results/eval-runner.ts b/apps/cli/src/commands/results/eval-runner.ts
index 9d5053cdd..91dd8386c 100644
--- a/apps/cli/src/commands/results/eval-runner.ts
+++ b/apps/cli/src/commands/results/eval-runner.ts
@@ -279,6 +279,8 @@ export function registerEvalRoutes(
         cwd,
         stdio: ['ignore', 'pipe', 'pipe'],
         env: { ...process.env },
+        // Windows requires shell:true to execute .cmd/.bat wrappers (e.g. npm-installed agentv.cmd)
+        shell: process.platform === 'win32',
       });
 
       run.process = child;
@@ -440,6 +442,7 @@ export function registerEvalRoutes(
         cwd,
         stdio: ['ignore', 'pipe', 'pipe'],
         env: { ...process.env },
+        shell: process.platform === 'win32',
       });
 
       run.process = child;
diff --git a/packages/core/test/evaluation/content-preprocessor.test.ts b/packages/core/test/evaluation/content-preprocessor.test.ts
index 22c562feb..bc0f4e049 100644
--- a/packages/core/test/evaluation/content-preprocessor.test.ts
+++ b/packages/core/test/evaluation/content-preprocessor.test.ts
@@ -42,8 +42,9 @@ describe('content preprocessors', () => {
     await writeFile(
       scriptPath,
       `const fs = require('node:fs');
+const path = require('node:path');
 const payload = JSON.parse(fs.readFileSync(0, 'utf8'));
-console.log('sheet:' + payload.original_path.split('/').pop());`,
+console.log('sheet:' + path.basename(payload.original_path));`,
       'utf8',
     );
 
diff --git a/packages/core/test/evaluation/llm-grader-multimodal.test.ts b/packages/core/test/evaluation/llm-grader-multimodal.test.ts
index bbf6e8c04..158d10245 100644
--- a/packages/core/test/evaluation/llm-grader-multimodal.test.ts
+++ b/packages/core/test/evaluation/llm-grader-multimodal.test.ts
@@ -374,8 +374,9 @@ describe('LlmGraderEvaluator multimodal', () => {
     await writeFile(
       scriptPath,
       `const fs = require('node:fs');
+const path = require('node:path');
 const payload = JSON.parse(fs.readFileSync(0, 'utf8'));
-console.log('spreadsheet:' + payload.original_path.split('/').pop());`,
+console.log('spreadsheet:' + path.basename(payload.original_path));`,
       'utf8',
     );
 
diff --git a/packages/core/test/evaluation/loaders/evaluator-parser.test.ts b/packages/core/test/evaluation/loaders/evaluator-parser.test.ts
index 5d7b393f0..bf2998c58 100644
--- a/packages/core/test/evaluation/loaders/evaluator-parser.test.ts
+++ b/packages/core/test/evaluation/loaders/evaluator-parser.test.ts
@@ -1571,7 +1571,7 @@ assertions:
         [tempDir],
         'test-5',
       ),
-    ).rejects.toThrow(/\.agentv\/templates\/missing-template\.yaml/);
+    ).rejects.toThrow(/\.agentv[/\\]templates[/\\]missing-template\.yaml/);
   });
 });