test(style[format]): Normalize bookmark tests

why: Ruff's format gate rejected the focused bookmark tests after the
stack was replayed onto the updated identity branch.

what:
- Apply Ruff's canonical wrapping to CLI and command tests.

Author: tony

tests/test_bookmark_cli.py

@@ -219,8 +219,7 @@ def test_bookmark_cli_human_list_output(
     assert agentgrep.main(["bookmark", "list"]) == 0
 
     assert capsys.readouterr().out == (
-        f"thread\t{THREAD_ID}\t-\t{CREATED_AT}\n"
-        f"record\t{RECORD_ID}\t{CONTENT_ID}\t{CREATED_AT}\n"
+        f"thread\t{THREAD_ID}\t-\t{CREATED_AT}\nrecord\t{RECORD_ID}\t{CONTENT_ID}\t{CREATED_AT}\n"
     )
 
 

tests/test_bookmark_commands.py

@@ -37,9 +37,7 @@ def test_bookmark_commands_are_hud_extensions() -> None:
     assert by_name["bookmark"].argument_hint == "[record|thread|content]"
     assert by_name["bookmark"].accepts_args is True
     assert by_name["bookmarks"].argument_hint == ""
-    assert commands.command_menu_label(by_name["bookmark"]) == (
-        "bookmark [record|thread|content]"
-    )
+    assert commands.command_menu_label(by_name["bookmark"]) == ("bookmark [record|thread|content]")
     assert commands.command_menu_label(by_name["bookmarks"]) == "bookmarks"