--cell-history

The commits that changed one cell of a file kept in a Git repository, newest first. The row is followed by a key, so a re-sorted file is not reported as every row changing, and renames are followed. When the file on disk differs from the last commit, its current state comes first as (not committed).

octa --cell-history prices.csv --history-column price --history-key id --history-value 2
octa --cell-history prices.csv --history-column price --history-row 5

Flags

Flag	Required?	Description
--cell-history	yes	The file. Must be inside a Git repository.
--history-column	yes	Column of the cell.
--history-key	one of	Key column(s) that identify the row, comma-separated.
--history-value	with key	The row's value in each key column, comma-separated, same order.
--history-row	one of	1-based row number, when there is no key.
--history-depth	no	How many commits to read, newest first. Default 50.

Output

commit   date              author  subject      value  change
a1b2c3d  2026-09-20 14:02  Tess    raise price  25     changed
9f8e7d6  2026-09-18 09:40  Tess    create       20     earliest

change is one of changed, row_added, row_removed, column_added, column_removed, or earliest for the oldest version read. On stderr: the versions where the key was not unique or missing (the row was matched by its position there), versions that could not be read, and a note when older commits exist beyond --history-depth.

See Cell History.

See also

• MCP: cell_history