diff --git a/Tools/autotest/autotest.py b/Tools/autotest/autotest.py
index 1c6d650756232..59163af331aa3 100755
--- a/Tools/autotest/autotest.py
+++ b/Tools/autotest/autotest.py
@@ -459,6 +459,7 @@ def run_step(step):
         return util.build_replay(board='SITL')
 
     if vehicle_binary is not None:
+        print(f"::group::{step}")
         try:
             binary = binary_path(step, debug=opts.debug)
             os.unlink(binary)
@@ -719,16 +720,17 @@ def run_tests(steps):
     passed = True
     failed = []
     failed_testinstances = dict()
-    should_close_group = False
+    print("::endgroup::")
+
     for step in steps:
-        if "build" in step:
-            should_close_group = True
         util.pexpect_close_all()
 
         t1 = time.time()
         print(">>>> RUNNING STEP: %s at %s" % (step, time.asctime()))
         try:
             success = run_step(step)
+            if "build" in step:
+                print("::endgroup::")
             testinstance = None
             if isinstance(success, tuple):
                 (success, testinstance) = success
@@ -736,9 +738,6 @@ def run_tests(steps):
                 results.add(step, '<span class="passed-text">PASSED</span>',
                             time.time() - t1)
                 print(">>>> PASSED STEP: %s at %s" % (step, time.asctime()))
-                if should_close_group:
-                    print("::endgroup::")
-                    should_close_group = False
             else:
                 print(">>>> FAILED STEP: %s at %s" % (step, time.asctime()))
                 passed = False
@@ -752,9 +751,9 @@ def run_tests(steps):
         except Exception as msg:
             passed = False
             failed.append(step)
+            print("::endgroup::")
             print(">>>> FAILED STEP: %s at %s (%s)" %
                   (step, time.asctime(), msg))
-            print(f"::error file=autotest.py::{msg}")
             traceback.print_exc(file=sys.stdout)
             results.add(step,
                         '<span class="failed-text">FAILED</span>',
diff --git a/Tools/autotest/vehicle_test_suite.py b/Tools/autotest/vehicle_test_suite.py
index cffa714ce9949..0687e13683234 100644
--- a/Tools/autotest/vehicle_test_suite.py
+++ b/Tools/autotest/vehicle_test_suite.py
@@ -9052,6 +9052,7 @@ def run_one_test_attempt(self, test, interact=False, attempt=1, suppress_stdout=
         else:
             self.progress('FAILED: "%s": %s (see %s)' %
                           (prettyname, repr(ex), test_output_filename))
+            print(f"::error::FAILED: {prettyname}: {repr(ex)} (see {test_output_filename})", file=sys.stdout)
             result.exception = ex
             result.debug_filename = test_output_filename
             if interact:
@@ -12301,7 +12302,9 @@ def run_tests(self, tests) -> List[Result]:
 
             for test in tests:
                 self.drain_mav_unparsed()
+                print(f"::group::{test.name}")
                 result_list.append(self.run_one_test(test))
+                print("::endgroup::")
 
         except pexpect.TIMEOUT:
             self.progress("Failed with timeout")
@@ -12312,6 +12315,7 @@ def run_tests(self, tests) -> List[Result]:
             if self.logs_dir:
                 if glob.glob("core*") or glob.glob("ap-*.core"):
                     self.check_logs("FRAMEWORK")
+            print("::endgroup::")
 
         if self.rc_thread is not None:
             self.progress("Joining RC thread")