diff --git a/tests/test_noema_document_review_context.py b/tests/test_noema_document_review_context.py
index e6ec2e6d70..063c95d08d 100644
--- a/tests/test_noema_document_review_context.py
+++ b/tests/test_noema_document_review_context.py
@@ -6,6 +6,9 @@
 import io
 import json
 import os
+import re
+import runpy
+import sys
 import zipfile
 from pathlib import Path
 
@@ -48,6 +51,7 @@ def _docx_entity_bytes() -> bytes:
 
 
 def _pr() -> dict[str, object]:
+    """Build a minimal PR payload shared by review-context tests."""
     return {
         "headRefOid": "head",
         "baseRefOid": "base",
@@ -105,6 +109,7 @@ def test_docx_text_reaches_the_actual_reviewer_payload(monkeypatch):
     encoded = base64.b64encode(raw).decode("ascii")
 
     def fake_run(args, stdin=None):
+        """Return base64-encoded DOCX bytes for the requested content ref."""
         assert "contents/docs/review.docx?ref=head" in args[2]
         return encoded
 
@@ -123,20 +128,26 @@ def fake_run(args, stdin=None):
     captured: dict[str, object] = {}
 
     class Response:
+        """Fake urllib response object supporting the context-manager protocol."""
         def __enter__(self):
+            """Enter the fake response context manager."""
             return self
 
         def __exit__(self, *args):
+            """Exit the fake response context manager."""
             return False
 
         def read(self):
+            """Return a fake chat-completion payload as JSON bytes."""
             verdict = {"decision": "comment", "summary": "checked", "findings": []}
             return json.dumps(
                 {"choices": [{"message": {"content": json.dumps(verdict)}}]}
             ).encode()
 
     class Opener:
+        """Fake urllib opener capturing the outgoing request payload."""
         def open(self, request):
+            """Capture the request body and return a fake response."""
             captured.update(json.loads(request.data.decode()))
             return Response()
 
@@ -200,6 +211,7 @@ def test_hwp_reader_contract_is_local_and_fail_closed(monkeypatch):
         raise AssertionError("expected failed local HWP reader to fail closed")
 
     def timed_out(*args, **kwargs):
+        """Simulate a hung reviewed HWP reader subprocess."""
         raise document.subprocess.TimeoutExpired(args[0], kwargs["timeout"])
 
     monkeypatch.setattr(document.subprocess, "run", timed_out)
@@ -243,20 +255,26 @@ def test_real_hwp_mcp_fixture_text_reaches_reviewer_payload(
     captured: dict[str, object] = {}
 
     class Response:
+        """Fake urllib response object supporting the context-manager protocol."""
         def __enter__(self):
+            """Enter the fake response context manager."""
             return self
 
         def __exit__(self, *args):
+            """Exit the fake response context manager."""
             return False
 
         def read(self):
+            """Return a fake chat-completion payload as JSON bytes."""
             verdict = {"decision": "comment", "summary": "checked", "findings": []}
             return json.dumps(
                 {"choices": [{"message": {"content": json.dumps(verdict)}}]}
             ).encode()
 
     class Opener:
+        """Fake urllib opener capturing the outgoing request payload."""
         def open(self, request):
+            """Capture the request body and return a fake response."""
             captured.update(json.loads(request.data.decode()))
             return Response()
 
@@ -275,3 +293,228 @@ def open(self, request):
     assert expected_text in prompt
     if fixture_name == "simple.hwp":
         assert "| 이름 | 회사 |" in prompt
+
+
+def test_oversize_raw_bytes_are_rejected_before_any_parsing():
+    """Raw input above the bounded 8 MiB cap fails closed without inspecting the suffix."""
+    raw = b"x" * (document.MAX_DOCUMENT_BYTES + 1)
+    with pytest.raises(
+        document.DocumentReadError,
+        match="document exceeds the bounded 8 MiB review input",
+    ):
+        document.extract_review_document("docs/huge.docx", raw)
+
+
+def test_unsupported_suffix_is_rejected():
+    """A file extension outside the supported set fails closed with the suffix named."""
+    with pytest.raises(
+        document.DocumentReadError,
+        match=r"unsupported review document format: \.txt",
+    ):
+        document.extract_review_document("docs/notes.txt", b"plain text")
+
+
+def _zip_bytes(entries: dict[str, bytes]) -> bytes:
+    """Build an in-memory ZIP archive from an arcname -> content mapping."""
+    output = io.BytesIO()
+    with zipfile.ZipFile(output, "w", zipfile.ZIP_DEFLATED) as archive:
+        for name, content in entries.items():
+            archive.writestr(name, content)
+    return output.getvalue()
+
+
+def test_docx_zip_with_too_many_entries_is_rejected():
+    """A DOCX archive with more than the bounded entry count fails closed."""
+    entries = {
+        f"part-{index}.xml": b""
+        for index in range(document.MAX_DOCUMENT_ZIP_ENTRIES + 1)
+    }
+    with pytest.raises(
+        document.DocumentReadError, match="DOCX archive has too many entries"
+    ):
+        document.extract_review_document("docs/many-entries.docx", _zip_bytes(entries))
+
+
+def test_docx_zip_over_the_bounded_unpacked_size_is_rejected(monkeypatch):
+    """A DOCX archive whose declared unpacked size exceeds the cap fails closed."""
+    monkeypatch.setattr(document, "MAX_DOCUMENT_ZIP_UNCOMPRESSED_BYTES", 5)
+    raw = _zip_bytes({"word/document.xml": b"0123456789"})
+    with pytest.raises(
+        document.DocumentReadError,
+        match="DOCX archive exceeds the bounded unpacked size",
+    ):
+        document.extract_review_document("docs/oversized-unpacked.docx", raw)
+
+
+def test_docx_missing_document_xml_is_rejected():
+    """A DOCX archive without word/document.xml fails closed with a clear message."""
+    raw = _zip_bytes({"word/other.xml": b"<x/>"})
+    with pytest.raises(
+        document.DocumentReadError, match=re.escape("DOCX archive has no word/document.xml")
+    ):
+        document.extract_review_document("docs/no-document-xml.docx", raw)
+
+
+def test_docx_document_xml_without_body_is_rejected():
+    """A document.xml with no w:body element fails closed."""
+    xml = """<?xml version="1.0" encoding="UTF-8"?>
+<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
+</w:document>"""
+    raw = _zip_bytes({"word/document.xml": xml.encode("utf-8")})
+    with pytest.raises(
+        document.DocumentReadError, match=re.escape("DOCX document.xml has no document body")
+    ):
+        document.extract_review_document("docs/no-body.docx", raw)
+
+
+def test_docx_with_only_empty_paragraphs_and_rowless_table_has_no_readable_text():
+    """Empty paragraphs and a table with no populated rows leave no readable text."""
+    xml = """<?xml version="1.0" encoding="UTF-8"?>
+<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
+  <w:body>
+    <w:p><w:pPr/></w:p>
+    <w:tbl><w:tr></w:tr></w:tbl>
+  </w:body>
+</w:document>"""
+    raw = _zip_bytes({"word/document.xml": xml.encode("utf-8")})
+    with pytest.raises(
+        document.DocumentReadError, match="DOCX contains no readable text"
+    ):
+        document.extract_review_document("docs/empty.docx", raw)
+
+
+def test_docx_paragraph_tabs_breaks_and_table_pipe_escaping():
+    """Tabs, line breaks, and a second table row exercise the paragraph/table branches."""
+    xml = """<?xml version="1.0" encoding="UTF-8"?>
+<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
+  <w:body>
+    <w:p><w:t>Section-Marker</w:t><w:tab/><w:t>After-Tab</w:t><w:br/><w:t>After-Break</w:t></w:p>
+    <w:tbl>
+      <w:tr><w:tc><w:p><w:t>a|b</w:t></w:p></w:tc><w:tc><w:p><w:t>c</w:t></w:p></w:tc></w:tr>
+      <w:tr></w:tr>
+    </w:tbl>
+    <w:sectPr/>
+  </w:body>
+</w:document>"""
+    raw = _zip_bytes({"word/document.xml": xml.encode("utf-8")})
+    text = document.extract_review_document("docs/tabs-breaks.docx", raw)
+    assert "Section-Marker\tAfter-Tab\nAfter-Break" in text
+    assert "a\\|b" in text
+
+
+def test_hwp_reader_env_unset_fails_closed(monkeypatch):
+    """HWP/HWPX extraction fails closed when the reviewed reader source is unset."""
+    monkeypatch.delenv(document.HWP_READER_ENV, raising=False)
+    with pytest.raises(
+        document.DocumentReadError,
+        match="reviewed hwp-mcp/rhwp reader is not configured",
+    ):
+        document.extract_review_document("docs/unconfigured.hwp", b"binary")
+
+
+def test_hwp_reader_subprocess_os_error_fails_closed(monkeypatch):
+    """An OSError starting the reviewed reader subprocess fails closed."""
+    monkeypatch.setenv(document.HWP_READER_ENV, "/trusted/hwp-mcp-source")
+
+    def raise_os_error(*_args, **_kwargs):
+        """Simulate a reviewed HWP reader subprocess that cannot start."""
+        raise OSError("node executable not found")
+
+    monkeypatch.setattr(document.subprocess, "run", raise_os_error)
+    with pytest.raises(
+        document.DocumentReadError,
+        match="reviewed hwp-mcp/rhwp reader could not start",
+    ):
+        document.extract_review_document("docs/no-node.hwp", b"binary")
+
+
+def test_hwp_reader_output_over_bounded_size_fails_closed(monkeypatch):
+    """Reader stdout larger than the bounded text cap fails closed."""
+    monkeypatch.setenv(document.HWP_READER_ENV, "/trusted/hwp-mcp-source")
+    completed = document.subprocess.CompletedProcess(
+        ["node"], 0, stdout=b"a" * (document.MAX_DOCUMENT_TEXT_BYTES + 1), stderr=b""
+    )
+    monkeypatch.setattr(document.subprocess, "run", lambda *args, **kwargs: completed)
+    with pytest.raises(
+        document.DocumentReadError,
+        match="reviewed hwp-mcp/rhwp reader exceeded the bounded output",
+    ):
+        document.extract_review_document("docs/too-long.hwp", b"binary")
+
+
+def test_hwp_reader_non_utf8_output_fails_closed(monkeypatch):
+    """Non-UTF-8 reader stdout fails closed instead of producing replacement text."""
+    monkeypatch.setenv(document.HWP_READER_ENV, "/trusted/hwp-mcp-source")
+    completed = document.subprocess.CompletedProcess(
+        ["node"], 0, stdout=b"\xff\xfe\xfa", stderr=b""
+    )
+    monkeypatch.setattr(document.subprocess, "run", lambda *args, **kwargs: completed)
+    with pytest.raises(
+        document.DocumentReadError,
+        match="reviewed hwp-mcp/rhwp reader returned non-UTF-8 text",
+    ):
+        document.extract_review_document("docs/bad-encoding.hwp", b"binary")
+
+
+def test_hwp_reader_empty_output_fails_closed(monkeypatch):
+    """Whitespace-only reader stdout fails closed as empty text."""
+    monkeypatch.setenv(document.HWP_READER_ENV, "/trusted/hwp-mcp-source")
+    completed = document.subprocess.CompletedProcess(
+        ["node"], 0, stdout=b"   \n\t  ", stderr=b""
+    )
+    monkeypatch.setattr(document.subprocess, "run", lambda *args, **kwargs: completed)
+    with pytest.raises(
+        document.DocumentReadError,
+        match="reviewed hwp-mcp/rhwp reader returned empty text",
+    ):
+        document.extract_review_document("docs/blank.hwp", b"binary")
+
+
+def test_bounded_text_truncates_multibyte_text_cleanly():
+    """Truncation of multibyte text stays valid UTF-8 and reports omitted bytes."""
+    text = "가" * ((document.MAX_DOCUMENT_TEXT_BYTES // 3) + 10)
+    bounded = document._bounded_text(text)
+    assert bounded != text
+    assert "[document text truncated;" in bounded
+    assert "bytes omitted]" in bounded
+    prefix = bounded.split("\n[document text truncated;", 1)[0]
+    assert len(prefix.encode("utf-8")) <= document.MAX_DOCUMENT_TEXT_BYTES
+
+
+def test_main_entrypoint_prints_text_for_a_valid_docx(tmp_path, monkeypatch, capsys):
+    """The __main__ CLI path prints extracted text and exits 0 for a valid DOCX."""
+    docx_path = tmp_path / "good.docx"
+    docx_path.write_bytes(_docx_bytes())
+    monkeypatch.setattr(sys, "argv", ["noema_review_document.py", str(docx_path)])
+
+    with pytest.raises(SystemExit) as exc_info:
+        runpy.run_path(
+            str(Path(__file__).resolve().parents[1] / "scripts/ci/noema_review_document.py"),
+            run_name="__main__",
+        )
+
+    assert exc_info.value.code == 0
+    assert "DOCX-REVIEW-MARKER" in capsys.readouterr().out
+
+
+def test_main_entrypoint_exits_1_for_a_missing_file(tmp_path, monkeypatch, capsys):
+    """The __main__ CLI path exits 1 and reports the error for a missing file."""
+    missing_path = tmp_path / "missing.docx"
+    monkeypatch.setattr(sys, "argv", ["noema_review_document.py", str(missing_path)])
+
+    with pytest.raises(SystemExit) as exc_info:
+        runpy.run_path(
+            str(Path(__file__).resolve().parents[1] / "scripts/ci/noema_review_document.py"),
+            run_name="__main__",
+        )
+
+    assert exc_info.value.code == 1
+    assert capsys.readouterr().err.strip() != ""
+
+
+def test_fetch_file_content_at_ref_rejects_malformed_base64(monkeypatch):
+    """A content response that is not valid base64 fails closed instead of decoding garbage."""
+    monkeypatch.setattr(noema, "run", lambda _args, stdin=None: "not*valid*base64!!")
+
+    with pytest.raises(RuntimeError, match="malformed base64"):
+        noema.fetch_file_content_at_ref("owner/repo", "docs/review.docx", "head")