diff --git a/.github/workflows/translation-check.yml b/.github/workflows/translation-check.yml
index ed8fab7f19..7ae834dc5b 100644
--- a/.github/workflows/translation-check.yml
+++ b/.github/workflows/translation-check.yml
@@ -5,13 +5,17 @@ on:
   pull_request:
     paths:
       - "src/translation/wininstaller/**"
+      - "src/translation/*.ts"                     # <-- ADD THIS
       - "tools/check-wininstaller-translations.sh"
+      - "tools/check-translations.py"              # <-- ADD THIS (Match your script name)
       - ".github/workflows/translation-check.yml"
   push:
     paths:
-      - "src/translation/wininstaller/**"
-      - "tools/check-wininstaller-translations.sh"
-      - ".github/workflows/translation-check.yml"
+      - 'src/translation/wininstaller/**'
+      - 'src/translation/*.ts'
+      - 'tools/check-wininstaller-translations.sh'
+      - 'tools/check-translations.py'              # <-- ADD THIS HERE TOO
+      - '.github/workflows/translation-check.yml'
 
 jobs:
   translation-check:
@@ -26,3 +30,5 @@ jobs:
         run: ./tools/check-wininstaller-translations.sh
       - name: "Check for duplicate hotkeys (will not fail)"
         run: perl ./tools/checkkeys.pl
+      - name: "Check application translations"
+        run: ./tools/check-translations.py --ts-dir src/translation
diff --git a/tools/check-translations.py b/tools/check-translations.py
new file mode 100755
index 0000000000..52ab5144ba
--- /dev/null
+++ b/tools/check-translations.py
@@ -0,0 +1,286 @@
+#!/usr/bin/env python3
+#
+##############################################################################
+# Copyright (c) 2026
+#
+# Author(s):
+#  ChatGPT
+#  ann0see
+#  JaminShanti
+#  Gemini
+#  The Jamulus Development Team
+#
+##############################################################################
+#
+# This program is free software; you can redistribute it and/or modify it under
+# the terms of the GNU General Public License as published by the Free Software
+# Foundation; either version 2 of the License, or (at your option) any later
+# version.
+#
+# This program is distributed in the hope that it will be useful, but WITHOUT
+# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
+# FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
+# details.
+#
+# You should have received a copy of the GNU General Public License along with
+# this program; if not, write to the Free Software Foundation, Inc.,
+# 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+#
+##############################################################################
+
+"""
+Qt TS translation checker.
+
+This tool validates Qt `.ts` translation files according to Qt Linguist
+semantics.
+Warnings are reported with best-effort line numbers. In strict mode, the
+presence of any warning results in a non-zero exit code to allow CI failure.
+"""
+
+import argparse
+import re
+import sys
+import xml.etree.ElementTree as ET
+from collections import defaultdict, Counter
+from dataclasses import dataclass
+from enum import IntEnum
+from pathlib import Path
+
+# Regex helpers
+PLACEHOLDER_RE = re.compile(r"%\d+")
+HTML_TAG_RE = re.compile(r"<[^>]+>")
+
+# ANSI escape codes
+BOLD = "\033[1m"
+CYAN = "\033[36m"
+YELLOW = "\033[33m"
+RED = "\033[31m"
+RESET = "\033[0m"
+
+
+# Severity Enum
+class Severity(IntEnum):
+    WARNING = 1
+    SEVERE = 2
+
+
+# Data structures
+@dataclass(frozen=True)
+class MessageContext:
+    ts_file: Path
+    line: int
+    lang: str
+    source: str
+    translation: str
+    tr_type: str
+    excerpt: str
+
+
+@dataclass(frozen=True)
+class WarningItem:
+    ts_file: Path
+    line: int
+    lang: str
+    message: str
+    severity: Severity
+
+
+# Helpers
+def approximate_message_lines(text: str):
+    """Yield approximate line numbers for <message> elements."""
+    lines = text.splitlines()
+    cursor = 0
+    for _ in range(text.count("<message")):
+        for i in range(cursor, len(lines)):
+            if "<message" in lines[i]:
+                cursor = i + 1
+                yield i + 1
+                break
+        else:
+            yield 0
+
+
+# Checks
+def check_language_header(ts_file: Path, root, file_lang: str):
+    header_lang = root.attrib.get("language", "")
+    if header_lang != file_lang:
+        return [WarningItem(ts_file, 0, file_lang,
+                            f"Language header mismatch '{header_lang}' != '{file_lang}'",
+                            Severity.WARNING)]
+    return []
+
+
+def check_empty_translation(ctx: MessageContext):
+    if not ctx.translation.strip() and ctx.tr_type != "unfinished":
+        return [WarningItem(ctx.ts_file, ctx.line, ctx.lang,
+                            f"Empty translation for '{ctx.excerpt}'",
+                            Severity.SEVERE)]
+    return []
+
+
+def check_placeholders(ctx: MessageContext):
+    if ctx.tr_type != "unfinished" and Counter(PLACEHOLDER_RE.findall(ctx.source)) != Counter(
+        PLACEHOLDER_RE.findall(ctx.translation)):
+        msg = (f"Placeholder mismatch for '{ctx.excerpt}'\n"
+               f"Source: {ctx.source}\n"
+               f"Trans:  {ctx.translation}")
+        return [WarningItem(ctx.ts_file, ctx.line, ctx.lang, msg, Severity.WARNING)]
+    return []
+
+
+def check_html(ctx: MessageContext):
+    if HTML_TAG_RE.search(ctx.source) and not HTML_TAG_RE.search(
+        ctx.translation) and ctx.tr_type != "unfinished":
+        msg = (f"HTML missing for '{ctx.excerpt}'\n"
+               f"Source: {ctx.source}\n"
+               f"Trans:  {ctx.translation}")
+        return [WarningItem(ctx.ts_file, ctx.line, ctx.lang, msg, Severity.WARNING)]
+    return []
+
+
+def check_whitespace(ctx: MessageContext):
+    if not ctx.translation or ctx.tr_type == "unfinished":
+        return []
+
+    # Check if leading/trailing whitespace presence matches between source and translation
+    src_lead = ctx.source != ctx.source.lstrip()
+    src_trail = ctx.source != ctx.source.rstrip()
+    tr_lead = ctx.translation != ctx.translation.lstrip()
+    tr_trail = ctx.translation != ctx.translation.rstrip()
+
+    if src_lead != tr_lead or src_trail != tr_trail:
+        return [WarningItem(ctx.ts_file, ctx.line, ctx.lang,
+                            f"Leading/trailing whitespace mismatch for '{ctx.excerpt}'",
+                            Severity.WARNING)]
+    return []
+
+
+def check_newline_consistency(ctx: MessageContext):
+    if ctx.source.endswith("\n") != ctx.translation.endswith("\n"):
+        return [WarningItem(ctx.ts_file, ctx.line, ctx.lang,
+                            f"Newline mismatch for '{ctx.excerpt}'",
+                            Severity.WARNING)]
+    return []
+
+
+# Detect warnings
+def detect_warnings(ts_file: Path, file_lang: str):
+    try:
+        text = ts_file.read_text(encoding="utf-8")
+        root = ET.fromstring(text)
+    except (OSError, ET.ParseError) as exc:
+        return [WarningItem(ts_file, 0, file_lang,
+                            f"Error reading or parsing XML: {exc}",
+                            Severity.SEVERE)]
+
+    warnings = []
+    warnings.extend(check_language_header(ts_file, root, file_lang))
+
+    message_lines = approximate_message_lines(text)
+
+    for context in root.findall("context"):
+        for message, line in zip(context.findall("message"), message_lines):
+
+            # Safely extract source text
+            source_elem = message.find("source")
+            source = "".join(source_elem.itertext()) if source_elem is not None else ""
+
+            # Safely extract translation text (handling Qt plural <numerusform> elements)
+            tr_elem = message.find("translation")
+            translation = ""
+            tr_type = ""
+            if tr_elem is not None:
+                tr_type = tr_elem.attrib.get("type", "")
+                numerus_forms = tr_elem.findall("numerusform")
+                if numerus_forms:
+                    translation = " ".join("".join(n.itertext()) for n in numerus_forms)
+                else:
+                    translation = "".join(tr_elem.itertext())
+
+            # Format a clean excerpt without blindly adding '...' to tiny strings
+            source_clean = source.strip().replace("\n", " ")
+            excerpt = source_clean[:30] + ("..." if len(source_clean) > 30 else "")
+
+            ctx = MessageContext(ts_file, line, file_lang, source, translation, tr_type, excerpt)
+
+            # All checks
+            warnings.extend(check_empty_translation(ctx))
+            warnings.extend(check_placeholders(ctx))
+            warnings.extend(check_html(ctx))
+            warnings.extend(check_whitespace(ctx))
+            warnings.extend(check_newline_consistency(ctx))
+
+    return warnings
+
+
+# CLI
+def main():
+    parser = argparse.ArgumentParser(description="Qt TS translation checker with extended rules")
+    parser.add_argument("--ts-dir", type=Path, default=Path("../src/translation"),
+                        help="Directory containing translation_*.ts files")
+    parser.add_argument("--strict", action="store_true",
+                        help="Exit non-zero if any warning is found")
+    args = parser.parse_args()
+
+    if not args.ts_dir.exists():
+        print(f"Directory not found: {args.ts_dir}", file=sys.stderr)
+        return 2
+
+    ts_files = sorted(args.ts_dir.glob("translation_*.ts"))
+    if not ts_files:
+        print(f"No TS files found in {args.ts_dir}", file=sys.stderr)
+        return 2
+
+    all_warnings = []
+    failures_by_language = defaultdict(lambda: {"severe": 0, "warning": 0})
+
+    for ts_file in ts_files:
+        lang = ts_file.stem.replace("translation_", "")
+        failures_by_language[lang]  # Initializes default counters
+        all_warnings.extend(detect_warnings(ts_file, lang))
+
+    # Group output by file
+    grouped = defaultdict(list)
+    for w in all_warnings:
+        grouped[w.ts_file].append(w)
+
+        if w.severity == Severity.SEVERE:
+            failures_by_language[w.lang]["severe"] += 1
+        else:
+            failures_by_language[w.lang]["warning"] += 1
+
+    # Detailed clean column output
+    for file in sorted(grouped.keys()):
+        messages = grouped[file]
+        print(f"\n{BOLD}File: {file.name}{RESET}")
+
+        for w in sorted(messages, key=lambda x: x.line):
+            color = RED if w.severity == Severity.SEVERE else YELLOW
+            sev_text = "SEVERE " if w.severity == Severity.SEVERE else "WARNING"
+
+            msg_lines = w.message.split("\n")
+            # Print the primary warning line
+            print(f"  {CYAN}Line {w.line:<4}{RESET} | {color}{sev_text}{RESET} | {msg_lines[0]}")
+
+            # Print any secondary data (like HTML/Placeholder source & trans contexts) properly aligned
+            for extra_line in msg_lines[1:]:
+                print(f"            |         | {extra_line}")
+
+    # Test summary
+    print("\n== Test Summary ==")
+    for lang in sorted(failures_by_language.keys()):
+        counts = failures_by_language[lang]
+        print(f"{BOLD}[{lang}]{RESET} Severe: {counts['severe']}, Warnings: {counts['warning']}")
+
+    total_severe = sum(f["severe"] for f in failures_by_language.values())
+    total_warning = sum(f["warning"] for f in failures_by_language.values())
+    print(f"\nTotal Severe: {total_severe}, Total Warnings: {total_warning}")
+
+    if total_severe > 0 or (args.strict and total_warning > 0):
+        return 1
+
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())