diff --git a/bookmark/catalogue.py b/bookmark/catalogue.py
new file mode 100644
index 0000000..f21a551
--- /dev/null
+++ b/bookmark/catalogue.py
@@ -0,0 +1,78 @@
+import csv
+import uuid
+from dataclasses import asdict, dataclass
+from pathlib import Path
+
+CATALOGUE_COLUMNS = [
+    "book_id",
+    "timestamp",
+    "size_bucket",
+    "front_wide_path",
+    "front_cropped_path",
+    "back_wide_path",
+    "back_cropped_path",
+    "title",
+    "author",
+    "isbn",
+    "match_confidence",
+    "match_status",
+    "status",
+    "checked_out_at",
+]
+
+
+@dataclass
+class BookRecord:
+    book_id: str
+    timestamp: str
+    size_bucket: str
+    front_wide_path: str = ""
+    front_cropped_path: str = ""
+    back_wide_path: str = ""
+    back_cropped_path: str = ""
+    title: str = ""
+    author: str = ""
+    isbn: str = ""
+    match_confidence: str = ""
+    match_status: str = ""
+    status: str = "captured"
+    checked_out_at: str = ""
+
+    def to_row(self) -> dict:
+        return asdict(self)
+
+
+class Catalogue:
+    def __init__(self, csv_path: Path) -> None:
+        self.csv_path = Path(csv_path)
+        if not self.csv_path.exists():
+            self.csv_path.parent.mkdir(parents=True, exist_ok=True)
+            with open(self.csv_path, "w", newline="") as f:
+                writer = csv.DictWriter(f, fieldnames=CATALOGUE_COLUMNS)
+                writer.writeheader()
+
+    def new_book_id(self) -> str:
+        return uuid.uuid4().hex[:12]
+
+    def append(self, record: BookRecord) -> None:
+        with open(self.csv_path, "a", newline="") as f:
+            writer = csv.DictWriter(f, fieldnames=CATALOGUE_COLUMNS)
+            writer.writerow(record.to_row())
+
+    def read_all(self) -> list[dict]:
+        with open(self.csv_path, newline="") as f:
+            return list(csv.DictReader(f))
+
+    def update_row(self, book_id: str, **fields_to_update) -> None:
+        rows = self.read_all()
+        found = False
+        for row in rows:
+            if row["book_id"] == book_id:
+                row.update({k: str(v) for k, v in fields_to_update.items()})
+                found = True
+        if not found:
+            raise KeyError(f"book_id {book_id} not found in catalogue")
+        with open(self.csv_path, "w", newline="") as f:
+            writer = csv.DictWriter(f, fieldnames=CATALOGUE_COLUMNS)
+            writer.writeheader()
+            writer.writerows(rows)
diff --git a/tests/test_catalogue.py b/tests/test_catalogue.py
new file mode 100644
index 0000000..3be939d
--- /dev/null
+++ b/tests/test_catalogue.py
@@ -0,0 +1,49 @@
+import pytest
+
+from bookmark.catalogue import Catalogue, BookRecord, CATALOGUE_COLUMNS
+
+
+def test_creating_catalogue_writes_header(tmp_path):
+    csv_path = tmp_path / "catalogue.csv"
+    Catalogue(csv_path)
+    assert csv_path.exists()
+    header_line = csv_path.read_text().splitlines()[0]
+    assert header_line.split(",") == CATALOGUE_COLUMNS
+
+
+def test_append_and_read_all_round_trip(tmp_path):
+    catalogue = Catalogue(tmp_path / "catalogue.csv")
+    record = BookRecord(book_id="abc123", timestamp="2026-07-13T00:00:00+00:00", size_bucket="medium")
+    catalogue.append(record)
+
+    rows = catalogue.read_all()
+
+    assert len(rows) == 1
+    assert rows[0]["book_id"] == "abc123"
+    assert rows[0]["size_bucket"] == "medium"
+    assert rows[0]["status"] == "captured"
+
+
+def test_new_book_id_is_unique(tmp_path):
+    catalogue = Catalogue(tmp_path / "catalogue.csv")
+    ids = {catalogue.new_book_id() for _ in range(100)}
+    assert len(ids) == 100
+
+
+def test_update_row_modifies_matching_row(tmp_path):
+    catalogue = Catalogue(tmp_path / "catalogue.csv")
+    catalogue.append(BookRecord(book_id="abc123", timestamp="t", size_bucket="small"))
+    catalogue.append(BookRecord(book_id="xyz789", timestamp="t", size_bucket="large"))
+
+    catalogue.update_row("abc123", title="Charlotte's Web", status="ready")
+
+    rows = {row["book_id"]: row for row in catalogue.read_all()}
+    assert rows["abc123"]["title"] == "Charlotte's Web"
+    assert rows["abc123"]["status"] == "ready"
+    assert rows["xyz789"]["title"] == ""
+
+
+def test_update_row_raises_for_unknown_book_id(tmp_path):
+    catalogue = Catalogue(tmp_path / "catalogue.csv")
+    with pytest.raises(KeyError):
+        catalogue.update_row("does-not-exist", title="x")