diff --git a/ci/build_test_OnCommit.groovy b/ci/build_test_OnCommit.groovy
index e68487d636..eaabf7fc02 100644
--- a/ci/build_test_OnCommit.groovy
+++ b/ci/build_test_OnCommit.groovy
@@ -210,7 +210,6 @@ pipeline {
                           windows.setup_bazel_remote_cache()
                           windows.install_dependencies()
                           windows.unit_test()
-                          windows.check_tests()
                         } finally {
                           windows.archive_test_artifacts()
                         }
diff --git a/ci/build_test_OnCommitWin.groovy b/ci/build_test_OnCommitWin.groovy
index 2ac7e697d3..337088d024 100644
--- a/ci/build_test_OnCommitWin.groovy
+++ b/ci/build_test_OnCommitWin.groovy
@@ -20,7 +20,6 @@ pipeline {
                           windows.clean()
                           windows.build()
                           windows.unit_test()
-                          windows.check_tests()
                           if (env.USE_BRANCH_NAME_IN_PACKAGE_NAME == "true") {
                             def safeBranchName = env.BRANCH_NAME.replaceAll('/', '_')
                             tag = "${safeBranchName}-${env.CUSTOM_TAG}"
diff --git a/ci/build_test_release.groovy b/ci/build_test_release.groovy
index 4f88ee17fb..bee5438fbf 100644
--- a/ci/build_test_release.groovy
+++ b/ci/build_test_release.groovy
@@ -24,7 +24,6 @@ pipeline {
                           windows.clean()
                           windows.build()
                           windows.unit_test()
-                          windows.check_tests()
                           def safeBranchName = env.BRANCH_NAME.replaceAll('/', '_')
                           def python_presence = ""
                           if (env.OVMS_PYTHON_ENABLED == "1") {
diff --git a/ci/loadWin.groovy b/ci/loadWin.groovy
index d2860b6fbc..983bbc03df 100644
--- a/ci/loadWin.groovy
+++ b/ci/loadWin.groovy
@@ -4,23 +4,23 @@ def cleanup_directories() {
     // First delete directories older than 14 days
     deleteOldDirectories()
     def command = 'ls c:\\Jenkins\\workspace | grep -oE "(PR-[0-9]*)$"'
-    def status = bat(returnStatus: true, script: command)
+    def status = bat(returnStatus: true, script: '@' + command)
     if ( status != 0) {
         error "Error: trying to list jenkins workspaces."
     }
-    def existing_workspace_string = bat(returnStatus: false, returnStdout: true, script: command)
+    def existing_workspace_string = bat(returnStatus: false, returnStdout: true, script: '@' + command)
 
     println existing_workspace_string
     def existing_workspace = existing_workspace_string.split(/\n/)
 
     command = 'ls c:\\ | grep -oE "(PR-[0-9]*)$"'
-    status = bat(returnStatus: true, script: command)
+    status = bat(returnStatus: true, script: '@' + command)
     if ( status != 0) {
         println "No PR-XXXX detected for cleanup."
         return
     }
 
-    def existing_prs_string = bat(returnStatus: false, returnStdout: true, script: command)
+    def existing_prs_string = bat(returnStatus: false, returnStdout: true, script: '@' + command)
 
     println existing_prs_string
     def existing_prs = existing_prs_string.split(/\n/)
@@ -46,7 +46,7 @@ def cleanup_directories() {
                 error "Error: trying to delete a directory that is not expected: " + pathToDelete
             } else {
                 println "Deleting: " + pathToDelete
-                status = bat(returnStatus: true, script: 'rmdir /s /q ' + pathToDelete)
+                status = bat(returnStatus: true, script: '@rmdir /s /q ' + pathToDelete)
                 if (status != 0) {
                     error "Error: Deleting directory ${pathToDelete} failed: ${status}. Check pipeline.log for details."
                 } else {
@@ -66,14 +66,14 @@ def get_short_bazel_path() {
 
 def deleteOldDirectories() {
     command = 'forfiles /P c:\\ /D -14 | grep -oE "(PR-[0-9]*)"'
-    status = bat(returnStatus: true, script: command)
+    status = bat(returnStatus: true, script: '@' + command)
     if ( status != 0) {
         println "No PR-XXXX older than 14 days for cleanup."
         return
     }
 
     // Check if directory was created more than 14 days ago
-    def existing_prs_string = bat(returnStatus: false, returnStdout: true, script: command)
+    def existing_prs_string = bat(returnStatus: false, returnStdout: true, script: '@' + command)
 
     println existing_prs_string
 
@@ -88,7 +88,7 @@ def deleteOldDirectories() {
             error "Error: trying to delete a directory that is not expected: " + pathToDelete
         } else {
             println "Deleting: " + pathToDelete
-            status = bat(returnStatus: true, script: 'rmdir /s /q ' + pathToDelete)
+            status = bat(returnStatus: true, script: '@rmdir /s /q ' + pathToDelete)
             if (status != 0) {
                 error "Error: Deleting directory ${pathToDelete} failed: ${status}. Check pipeline.log for details."
             } else {
@@ -100,7 +100,7 @@ def deleteOldDirectories() {
 
 def install_dependencies() {
     println "Install dependencies on node: NODE_NAME = ${env.NODE_NAME}"
-    def status = bat(returnStatus: true, script: 'windows_install_build_dependencies.bat ' + get_short_bazel_path() + ' ' + env.OVMS_CLEAN_EXPUNGE)
+    def status = bat(returnStatus: true, script: '@windows_install_build_dependencies.bat ' + get_short_bazel_path() + ' ' + env.OVMS_CLEAN_EXPUNGE)
     if (status != 0) {
         error "Error: Windows install dependencies failed: ${status}. Check pipeline.log for details."
     } else {
@@ -109,9 +109,9 @@ def install_dependencies() {
 }
 
 def clean() {
-    def output1 = bat(returnStdout: true, script: 'windows_clean_build.bat ' + get_short_bazel_path() + ' ' + env.OVMS_CLEAN_EXPUNGE)
+    def output1 = bat(returnStdout: true, script: '@windows_clean_build.bat ' + get_short_bazel_path() + ' ' + env.OVMS_CLEAN_EXPUNGE)
     if(fileExists('dist\\windows\\ovms')){
-        def status_del = bat(returnStatus: true, script: 'rmdir /s /q dist\\windows\\ovms')
+        def status_del = bat(returnStatus: true, script: '@rmdir /s /q dist\\windows\\ovms')
         if (status_del != 0) {
             error "Error: Deleting existing ovms directory failed ${status_del}. Check pipeline.log for details."
         } else {
@@ -123,21 +123,21 @@ def clean() {
 def build(){
     println "OVMS_PYTHON_ENABLED=${env.OVMS_PYTHON_ENABLED}"
     def pythonOption = env.OVMS_PYTHON_ENABLED == "0" ? "--no_python" : "--with_python"
-    def status = bat(returnStatus: true, script: 'windows_build.bat ' + get_short_bazel_path() + ' ' + pythonOption + ' --with_tests') 
-    status = bat(returnStatus: true, script: 'grep "Build completed successfully" win_build.log"')
+    def status = bat(returnStatus: true, script: '@windows_build.bat ' + get_short_bazel_path() + ' ' + pythonOption + ' --with_tests') 
+    status = bat(returnStatus: true, script: '@grep "Build completed successfully" win_build.log"')
     if (status != 0) {
         error "Error: Windows build failed ${status}. Check win_build.log for details."
     } else {
         echo "Build successful."
     }
-    def status_pkg = bat(returnStatus: true, script: 'windows_create_package.bat ' + get_short_bazel_path() + ' ' + pythonOption)
+    def status_pkg = bat(returnStatus: true, script: '@windows_create_package.bat ' + get_short_bazel_path() + ' ' + pythonOption)
     if (status_pkg != 0) {
         error "Error: Windows package failed ${status_pkg}."
     } else {
         echo "Windows package created successfully."
     }
     def unzipCmd = "tar -xf dist\\windows\\ovms.zip"
-    def status_unzip = bat(returnStatus: true, script: "${unzipCmd}")
+    def status_unzip = bat(returnStatus: true, script: '@' + "${unzipCmd}")
     if (status_unzip != 0) {
         error "Error: Unzipping package failed: ${status_unzip}."
     } else {
@@ -149,7 +149,7 @@ def clone_sdl_repo()
 {
     if(!fileExists('sdl_repo')){
         println "Starting code signing"
-        def statusPull = bat(returnStatus: true, script: 'git clone -b ' + env.SIGN_REPO_BRANCH + ' ' + env.SIGN_REPO + ' sdl_repo')
+        def statusPull = bat(returnStatus: true, script: '@git clone -b ' + env.SIGN_REPO_BRANCH + ' ' + env.SIGN_REPO + ' sdl_repo')
         if (statusPull != 0) {
             error "Error: Downloading sdl_repo failed ${statusPull}. Check pipeline.log for details."
         } else {
@@ -158,7 +158,7 @@ def clone_sdl_repo()
     }else{
         println "Pulling latest changes in sdl_repo"
         dir('sdl_repo') {
-            def statusPull = bat(returnStatus: true, script: 'git fetch && git reset --hard origin/'+env.SIGN_REPO_BRANCH)
+            def statusPull = bat(returnStatus: true, script: '@git fetch && git reset --hard origin/'+env.SIGN_REPO_BRANCH)
             if (statusPull != 0) {
                 error "Error: Pulling latest changes in sdl_repo failed ${statusPull}. Check pipeline.log for details."
             } else {
@@ -172,7 +172,7 @@ def clone_bdba_repo()
 {
     if(!fileExists('repo_ci_infra')){
         println "Starting BDBA infrastructure download"
-        def statusPull = bat(returnStatus: true, script: 'git clone -b ' + env.BDBA_REPO_BRANCH + ' ' + env.BDBA_REPO + ' repo_ci_infra')
+        def statusPull = bat(returnStatus: true, script: '@git clone -b ' + env.BDBA_REPO_BRANCH + ' ' + env.BDBA_REPO + ' repo_ci_infra')
         if (statusPull != 0) {
             error "Error: Downloading BDBA infrastructure failed ${statusPull}. Check pipeline.log for details."
         } else {
@@ -181,7 +181,7 @@ def clone_bdba_repo()
     }else{
         println "Pulling latest changes in BDBA infrastructure"
         dir('repo_ci_infra') {
-            def statusPull = bat(returnStatus: true, script: 'git fetch && git reset --hard origin/'+env.BDBA_REPO_BRANCH)
+            def statusPull = bat(returnStatus: true, script: '@git fetch && git reset --hard origin/'+env.BDBA_REPO_BRANCH)
             if (statusPull != 0) {
                 error "Error: Pulling latest changes in BDBA infrastructure failed ${statusPull}. Check pipeline.log for details."
             } else {
@@ -193,7 +193,7 @@ def clone_bdba_repo()
 
 def sign(){
     println "SIGNING_USER=${env.SIGNING_USER}"
-    def status = bat(returnStatus: true, script: 'ci\\windows_sign.bat ' + env.SIGNING_USER + ' dist\\windows ' + env.OVMS_PYTHON_ENABLED)
+    def status = bat(returnStatus: true, script: '@ci\\windows_sign.bat ' + env.SIGNING_USER + ' dist\\windows ' + env.OVMS_PYTHON_ENABLED)
     if (status != 0) {
         error "Error: Windows code signing failed ${status}. Check win_sign.log for details."
     } else {
@@ -203,7 +203,7 @@ def sign(){
 
 def bdba(){
     println "Starting BDBA scan"
-    def status = bat(returnStatus: true, script: 'ci\\windows_bdba.bat ' + env.BDBA_CREDS_PSW + ' dist\\windows sdl_repo\\ovms-package')
+    def status = bat(returnStatus: true, script: '@ci\\windows_bdba.bat ' + env.BDBA_CREDS_PSW + ' dist\\windows sdl_repo\\ovms-package')
     if (status != 0) {
         error "Error: Windows BDBA scan failed ${status}. Check win_bdba.log for details."
     } else {
@@ -214,7 +214,7 @@ def bdba(){
 def download_package(){
     println "Downloading package from URL: ${env.PACKAGE_URL}"
     if(!fileExists('dist\\windows')){
-        def status = bat(returnStatus: true, script: 'mkdir dist\\windows')
+        def status = bat(returnStatus: true, script: '@mkdir dist\\windows')
         if (status != 0) {
             error "Error: Creating dist\\windows directory failed ${status}. Check pipeline.log for details."
         } else {
@@ -223,7 +223,7 @@ def download_package(){
     }
     dir('dist\\windows') {
         if(fileExists('ovms.zip')){
-            def status_del = bat(returnStatus: true, script: 'del /f ovms.zip')
+            def status_del = bat(returnStatus: true, script: '@del /f ovms.zip')
             if (status_del != 0) {
                 error "Error: Deleting existing ovms.zip failed ${status_del}. Check pipeline.log for details."
             } else {
@@ -231,20 +231,20 @@ def download_package(){
             }
         }
         if(fileExists('ovms')){
-            def status_del = bat(returnStatus: true, script: 'rmdir /s /q ovms')
+            def status_del = bat(returnStatus: true, script: '@rmdir /s /q ovms')
             if (status_del != 0) {
                 error "Error: Deleting existing ovms directory failed ${status_del}. Check pipeline.log for details."
             } else {
                 echo "Existing ovms directory deleted successfully."
             }
         }
-        def status = bat(returnStatus: true, script: 'curl -L -k -o ovms.zip ' + env.PACKAGE_URL)
+        def status = bat(returnStatus: true, script: '@curl -L -k -o ovms.zip ' + env.PACKAGE_URL)
         if (status != 0) {
             error "Error: Downloading package failed ${status}. Check pipeline.log for details."
         } else {
             echo "Package downloaded successfully."
         }
-        def status_unzip = bat(returnStatus: true, script: 'tar -xf ovms.zip')
+        def status_unzip = bat(returnStatus: true, script: '@tar -xf ovms.zip')
         if (status_unzip != 0) {
             error "Error: Unzipping package failed: ${status_unzip}."
         } else {
@@ -256,44 +256,64 @@ def download_package(){
 def unit_test(){
     println "OVMS_PYTHON_ENABLED=${env.OVMS_PYTHON_ENABLED}"
     def pythonOption = env.OVMS_PYTHON_ENABLED == "0" ? "--no_python" : "--with_python"
-    status = bat(returnStatus: true, script: 'windows_test.bat ' + get_short_bazel_path() + ' ' + pythonOption)
-    if (status != 0) {
-        error "Error: Windows build test failed ${status}. Check win_build_test.log for details."
-    } else {
-        echo "Build successful."
-    }
-    status = bat(returnStatus: true, script: 'grep -A 4 bazel-bin/src/ovms_test.exe win_build_test.log | grep "Build completed successfully"')
+    boolean hasError = false
+    def errorReasons = []
+
+    def status = bat(returnStatus: true, script: '@windows_test.bat ' + get_short_bazel_path() + ' ' + pythonOption)
     if (status != 0) {
-        error "Error: Windows build test failed ${status}. Check win_build_test.log for details."
+        hasError = true
+        errorReasons << "Windows build test failed ${status}. Check win_build_test.log for details."
     } else {
         echo "Build successful."
     }
-}
 
-def check_tests(){
-    def status = bat(returnStatus: true, script: 'grep "       OK " win_test_summary.log')
+    status = bat(returnStatus: true, script: '@grep "       OK " win_test_summary.log')
     if (status != 0) {
-            error "Error: Windows run test failed ${status}. Expecting passed tests and no passed tests detected. Check win_test_summary.log for details."
+        hasError = true
+        errorReasons << "Windows run test failed ${status}. Expecting passed tests and no passed tests detected. Check win_test_summary.log for details."
     } else {
-        def passed = bat(returnStatus: false, returnStdout: true, script: 'grep "       OK " win_test_summary.log | wc -l')
-        echo "Success: Windows run test passed ${status}. ${passed} passed tests . Check win_test_summary.log for details."
+        def passed = bat(returnStatus: false, returnStdout: true, script: '@grep "       OK " win_test_summary.log | wc -l').trim()
+        echo "Detected ${passed} passed tests . Check win_test_summary.log for details."
     }
 
-    status = bat(returnStatus: true, script: 'grep "  FAILED  " win_test_summary.log')
+    status = bat(returnStatus: true, script: '@grep -a -E "^\\[  FAILED  \\].*[(][0-9]+ ms[)]$" win_full_test.log')
     if (status == 0) {
-            def failed = bat(returnStatus: false, returnStdout: true, script: 'grep "  FAILED  " win_test_summary.log | wc -l')
-            error "Error: Windows run test failed ${status}. ${failed} failed tests . Check win_test_summary.log for details."
+        hasError = true
+        def failed = bat(returnStatus: false, returnStdout: true, script: '@grep -a -E "^\\[  FAILED  \\].*[(][0-9]+ ms[)]$" win_full_test.log | awk "!seen[$0]++" | wc -l').trim()
+        def failedTestsList = bat(returnStatus: false, returnStdout: true, script: '@grep -a -E "^\\[  FAILED  \\].*[(][0-9]+ ms[)]$" win_full_test.log | awk "!seen[$0]++"').trim()
+        errorReasons << "Windows run test failed with ${failed} failed tests. Failed tests:\n${failedTestsList}\nCheck win_test_summary.log for details."
     } else {
-        echo "Run test no FAILED detected."
+        echo "no FAILED tests detected."
     }
 
-    status = bat(returnStatus: true, script: 'grep "  PASSED  " win_full_test.log')
+    // PASSED/crash detection is handled by windows_test.bat (via windows_parse_tests.bat).
+    // The bat exit code is the authoritative signal; win_test_summary.log contains diagnostics.
+    status = bat(returnStatus: true, script: '@grep -a "\\[  PASSED  \\] " win_full_test.log')
     if (status != 0) {
-            error "Error: Windows run test failed ${status}. Expecting   PASSED   at the end of log. Check pipeline.log for details."
+        hasError = true
+        def markerLine = bat(returnStatus: true, script: '@grep -n "Check tests summary in" win_test_summary.log | head -1')
+        def summaryContent
+        if (markerLine == 0) {
+            summaryContent = bat(returnStatus: false, returnStdout: true, script: '@grep -n "Check tests summary in" win_test_summary.log | head -1 | cut -d: -f1 | xargs -I{} head -{} win_test_summary.log').trim()
+        } else {
+            summaryContent = bat(returnStatus: false, returnStdout: true, script: '@head -150 win_test_summary.log').trim()
+        }
+        errorReasons << "Windows run test failed. PASSED summary not found in win_full_test.log.\n${summaryContent}"
     } else {
         echo "Success: Windows run test finished with success."
     }
 
+    status = bat(returnStatus: true, script: '@grep -A 4 bazel-bin/src/ovms_test.exe win_build_test.log | grep "Build completed successfully"')
+    if (status != 0) {
+        hasError = true
+        errorReasons << "Windows build test failed ${status}. Check win_build_test.log for details."
+    } else {
+        echo "Build successful."
+    }
+
+    if (hasError) {
+        error "Error: unit_test() detected failures:\n- " + errorReasons.join("\n- ")
+    }
 }
 
 // Post build steps
@@ -329,13 +349,13 @@ def setup_bazel_remote_cache(){
     def content = "build --remote_cache=\"${bazel_remote_cache_url}\""
     def filePath = '.user.bazelrc'
     def command = "echo ${content} > ${filePath}"
-    status = bat(returnStatus: true, script: command)
+    status = bat(returnStatus: true, script: '@' + command)
     if ( status != 0) {
         println "Failed to set up bazel remote cache for Windows"
         return
     }
     command = "cat ${filePath}"
-    status = bat(returnStatus: true, script: command)
+    status = bat(returnStatus: true, script: '@' + command)
     if ( status != 0) {
         println "Failed to read file"
         return
diff --git a/spelling-whitelist.txt b/spelling-whitelist.txt
index 85763a1742..4bac06c568 100644
--- a/spelling-whitelist.txt
+++ b/spelling-whitelist.txt
@@ -29,4 +29,11 @@ demos/vlm_npu/README.md:157: mane ==> main, many, maine
 demos/vlm_npu/README.md:218: mane ==> main, many, maine
 demos/integration_with_OpenWebUI/README.md:423: Buildin ==> Building, Build in
 src/test/llm/output_parsers/lfm2_output_parser_test.cpp
+windows_parse_tests.bat:35: seh ==> she
+windows_parse_tests.bat:119: SEH ==> SHE
+windows_parse_tests.bat:123: SEH ==> SHE
+windows_parse_tests.bat:134: SEH ==> SHE
+windows_parse_tests.bat:136: SEH ==> SHE
+windows_parse_tests.bat:141: SEH ==> SHE
+windows_parse_tests.bat:144: SEH ==> SHE
 src/test/llm/output_parsers/gemma4_output_parser_test.cpp
diff --git a/src/test/schema_test.cpp b/src/test/schema_test.cpp
index c97d6cfe2e..2b5aaf11fd 100644
--- a/src/test/schema_test.cpp
+++ b/src/test/schema_test.cpp
@@ -1155,7 +1155,6 @@ TEST(SchemaTest, ModelConfigPluginConfigPositive) {
         }
     ]
     })";
-
     rapidjson::Document modelConfigSeqNegativeDoc;
     modelConfigSeqNegativeDoc.Parse(modelConfigTimeoutNegative);
     auto result = ovms::validateJsonAgainstSchema(modelConfigSeqNegativeDoc, ovms::MODELS_CONFIG_SCHEMA.c_str());
diff --git a/windows_install_build_dependencies.bat b/windows_install_build_dependencies.bat
index 7f32d6143c..91df6decff 100644
--- a/windows_install_build_dependencies.bat
+++ b/windows_install_build_dependencies.bat
@@ -466,7 +466,7 @@ if !errorlevel! neq 0 exit /b !errorlevel!
 %python_path%\python.exe -m pip install --upgrade pip
 if !errorlevel! neq 0 exit /b !errorlevel!
 :: setuptools<60.0 required for numpy1.23 on python311 to install
-%python_path%\python.exe -m pip install "numpy==2.2.5" "Jinja2==3.1.6" "MarkupSafe==3.0.2"
+%python_path%\python.exe -m pip install "numpy==2.2.5" "Jinja2==3.1.6" "MarkupSafe==3.0.2" "pytest==8.3.5"
 if !errorlevel! neq 0 exit /b !errorlevel!
 echo [INFO] Python %python_version% installed: %python_path%
 goto install_curl
diff --git a/windows_parse_tests.bat b/windows_parse_tests.bat
new file mode 100644
index 0000000000..c4b28e3658
--- /dev/null
+++ b/windows_parse_tests.bat
@@ -0,0 +1,168 @@
+::
+:: Copyright (c) 2026 Intel Corporation
+::
+:: Licensed under the Apache License, Version 2.0 (the "License");
+:: you may not use this file except in compliance with the License.
+:: You may obtain a copy of the License at
+::
+::      http:::www.apache.org/licenses/LICENSE-2.0
+::
+:: Unless required by applicable law or agreed to in writing, software
+:: distributed under the License is distributed on an "AS IS" BASIS,
+:: WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+:: See the License for the specific language governing permissions and
+:: limitations under the License.
+::
+@echo off
+setlocal EnableExtensions EnableDelayedExpansion
+
+set "fullLog=%~1"
+set "summaryLog=%~2"
+if not defined fullLog set "fullLog=win_full_test.log"
+if not defined summaryLog set "summaryLog=win_test_summary.log"
+
+if not exist "!fullLog!" (
+    echo [ERROR] Full test log not found: !fullLog!
+    exit /b 1
+)
+if not exist "!summaryLog!" (
+    echo [WARN] Summary log not found: !summaryLog!
+)
+
+set "parserOutputTmp=!summaryLog!.parse.tmp"
+set "summaryBackupTmp=!summaryLog!.orig.tmp"
+
+set "CRASH_PATTERN=segmentation fault\|segfault\|abnormal termination\|access violation\|sigsegv\|seh exception\|0xc0000005\|unknown file: error:"
+set "FAILED_TEST_PATTERN=^\[  FAILED  \].*([0-9][0-9]* ms)$"
+
+:: Check for per-test FAILED markers first - do not allow PASSED text to mask test failures
+:: Match only timed gtest FAILED lines to avoid duplicates from the final "listed below" section.
+grep -a -q "%FAILED_TEST_PATTERN%" "%fullLog%"
+if !errorlevel! equ 0 goto :exit_build_error
+
+:: Also check for segmentation faults or crashes
+grep -a -q -i "%CRASH_PATTERN%" "%fullLog%"
+if !errorlevel! equ 0 goto :exit_build_error
+
+:: Consider the run successful only if the global PASSED summary is present anywhere in the log
+:: (it can appear hundreds of lines before EOF due to SKIPPED and debug trace lines following it)
+grep -a -q "\[  PASSED  \] [0-9]" "%fullLog%"
+if !errorlevel! equ 0 exit /b 0
+
+:: If we reach here, tests did not complete correctly (no FAILED/crash marker but no final PASSED summary)
+goto :exit_build_error
+
+:exit_build_error
+if exist "%parserOutputTmp%" del /f /q "%parserOutputTmp%"
+if exist "%summaryBackupTmp%" del /f /q "%summaryBackupTmp%"
+
+set "segfaultDetected=0"
+grep -a -q -i "%CRASH_PATTERN%" "!fullLog!"
+if !errorlevel! equ 0 set "segfaultDetected=1"
+
+(
+echo.
+echo [ERROR] FAILED TESTS OR CRASHES DETECTED:
+echo.
+echo === Failed Tests ^(from summary/full log^) ===
+grep -a "!FAILED_TEST_PATTERN!" "!fullLog!" | awk "!seen[$0]++"
+echo.
+echo === Last Successful Test ===
+grep -a " OK ]" "!fullLog!" | tail -1
+echo.
+if "!segfaultDetected!"=="1" (
+    echo === Last Running Test ^(likely the one that failed^) ===
+    set "lastRunEntry="
+    for /F "delims=" %%A in ('grep -a "\[ RUN" "!fullLog!" ^| tail -1') do (
+        set "lastRunEntry=%%A"
+    )
+    if defined lastRunEntry (
+        echo !lastRunEntry!
+    ) else (
+        echo [WARN] No gtest RUN marker found in !fullLog!.
+    )
+    echo.
+    echo === Output from Last Running Test to End of Log ===
+    set "lastRunLine="
+    for /F "tokens=1 delims=:" %%A in ('grep -a -n "\[ RUN" "!fullLog!" ^| tail -1') do (
+        set "lastRunLine=%%A"
+    )
+    if defined lastRunLine (
+        sed -n "!lastRunLine!,$p" "!fullLog!" | head -120
+    ) else (
+        echo [WARN] Could not determine last RUN line. Showing recent RUN markers and log tail.
+        grep -a -n "\[ RUN" "!fullLog!" | tail -3
+        echo.
+        tail -20 "!fullLog!"
+    )
+    echo.
+)
+echo === Context Around First FAILED Test ===
+set "firstFailedLine="
+set "firstFailedRunLine="
+for /F "tokens=1 delims=:" %%A in ('grep -a -n "!FAILED_TEST_PATTERN!" "!fullLog!" ^| head -1') do (
+    set "firstFailedLine=%%A"
+)
+if defined firstFailedLine (
+    for /F "tokens=1 delims=:" %%B in ('sed -n "1,!firstFailedLine!p" "!fullLog!" ^| grep -a -n "\[ RUN" ^| tail -1') do (
+        set "firstFailedRunLine=%%B"
+    )
+    if defined firstFailedRunLine (
+        sed -n "!firstFailedRunLine!,$p" "!fullLog!" | head -160
+    ) else (
+        echo [WARN] Could not determine RUN line for first FAILED test.
+    )
+) else (
+    echo [INFO] No per-test FAILED entry with timing found.
+)
+echo.
+echo === SEH/Access Violation Context ===
+set "firstSehLine="
+set "firstSehRunLine="
+set "firstSehRunEntry="
+for /F "tokens=1 delims=:" %%A in ('grep -a -n -i "unknown file: error: SEH exception\|0xc0000005\|access violation\|SEH exception" "!fullLog!" ^| head -1') do (
+    set "firstSehLine=%%A"
+)
+if defined firstSehLine (
+    for /F "tokens=1 delims=:" %%B in ('sed -n "1,!firstSehLine!p" "!fullLog!" ^| grep -a -n "\[ RUN" ^| tail -1') do (
+        set "firstSehRunLine=%%B"
+    )
+    for /F "delims=" %%C in ('sed -n "1,!firstSehLine!p" "!fullLog!" ^| grep -a "\[ RUN" ^| tail -1') do (
+        set "firstSehRunEntry=%%C"
+    )
+    if defined firstSehRunEntry (
+        echo [INFO] Unit test at first SEH marker: !firstSehRunEntry!
+    ) else (
+        echo [WARN] Could not determine unit test name at first SEH marker.
+    )
+    if defined firstSehRunLine (
+        sed -n "!firstSehRunLine!,$p" "!fullLog!" | head -160
+    ) else (
+        echo [WARN] Could not determine RUN line for SEH exception entry.
+    )
+) else (
+    echo [INFO] No SEH/Access Violation entry found.
+)
+echo.
+echo === Segfault/Crash Messages ^(if any^) ===
+grep -a -i "%CRASH_PATTERN%\|stack trace" "!fullLog!" || echo ^(none found^)
+echo.
+echo [ERROR] Check tests summary in '!summaryLog!' and tests logs in '!fullLog!'. Rerun failed test with: windows_setupvars.bat and %cd%\bazel-bin\src\ovms_test.exe --gtest_filter='*.*'
+) > "!parserOutputTmp!" 2>&1
+
+if exist "!summaryLog!" (
+    copy /Y "!summaryLog!" "!summaryBackupTmp!" > nul
+) else (
+    type nul > "!summaryBackupTmp!"
+)
+
+(
+    type "%parserOutputTmp%"
+    echo.
+    type "%summaryBackupTmp%"
+) > "%summaryLog%"
+
+if exist "!parserOutputTmp!" del /f /q "!parserOutputTmp!"
+if exist "!summaryBackupTmp!" del /f /q "!summaryBackupTmp!"
+
+exit /b 1
diff --git a/windows_test.bat b/windows_test.bat
index dfd1a8c424..e6b5ca965d 100644
--- a/windows_test.bat
+++ b/windows_test.bat
@@ -98,10 +98,12 @@ if !errorlevel! neq 0 exit /b !errorlevel!
 
 :: Run install_ovms_service.bat unit tests
 echo Running install_ovms_service.bat unit tests...
-python -m pytest tests\python\test_install_ovms_service_windows.py -v 2>&1 | tee win_install_service_test.log
-if !errorlevel! neq 0 (
+python -m pytest tests\python\test_install_ovms_service_windows.py -v > win_install_service_test.log 2>&1
+set "pytestExitCode=!errorlevel!"
+type win_install_service_test.log
+if !pytestExitCode! neq 0 (
     echo [ERROR] install_ovms_service.bat unit tests failed. See win_install_service_test.log.
-    exit /b !errorlevel!
+    exit /b !pytestExitCode!
 )
 echo [INFO] install_ovms_service.bat unit tests passed.
 
@@ -113,13 +115,34 @@ echo Running: %runTest%
 set regex="\[  .* ms"
 set sed_clean="s/ (.* ms)//g"
 C:\Windows\System32\tar.exe -a -c -f win_test_log.zip win_full_test.log
-grep -a %regex% win_full_test.log | sed %sed_clean% > win_test_summary.log
-grep -a %regex% win_full_test.log | sed %sed_clean% | grep -q " FAILED "
-if !errorlevel! equ 0 goto :exit_build_error
-:exit_build 
+
+:: Create summary log with filtered results, always create the file even if grep finds no matches
+ grep -a %regex% win_full_test.log > win_test_summary.tmp
+ if !errorlevel! equ 0 (
+     sed %sed_clean% win_test_summary.tmp > win_test_summary.log 2>&1
+ ) else (
+     echo No matching test results found > win_test_summary.log
+ )
+ if exist win_test_summary.tmp del /f /q win_test_summary.tmp
+
+:: Parse logs and decide final test status using dedicated parser script
+:: Skip expensive parsing only if PASSED summary exists AND no FAILED markers
+grep -a -q "\[  PASSED  \] " win_full_test.log
+set "hasPassed=!errorlevel!"
+grep -a -q "\[  FAILED  \] " win_full_test.log
+set "hasFailed=!errorlevel!"
+if !hasPassed! equ 0 if !hasFailed! neq 0 (
+    echo [INFO] Tests finished with no failures. Check the summary in win_test_summary.log.
+    exit /b 0
+)
+call %cd%\windows_parse_tests.bat win_full_test.log win_test_summary.log
+set "parseExitCode=!errorlevel!"
+if !parseExitCode! neq 0 exit /b !parseExitCode!
+
 echo [INFO] Tests finished with no failures. Check the summary in win_test_summary.log.
 exit /b 0
+
 :exit_build_error
-echo [ERROR] Check tests summary in 'win_test_summary.log' and tests logs in 'win_full_test.log'. Rerun failed test with: windows_setupvars.bat and %cd%\bazel-bin\src\ovms_test.exe --gtest_filter='*.*'
+echo [ERROR] windows_test.bat failed before test parsing stage.
 exit /b 1
 endlocal