Skip to main content
DocumentationLibraryReference
On this page

Kater CLI Pages you have open or were sent

kater answer

Reads the rows that a chart, table or panel on one of your Kater pages shows.

Give it the answer id that kater here lists, or a run’s snapshot_id. Nothing runs on the warehouse.

Examples

Terminal window
# List your tabs, and note the answer id beside the chart you want
kater here
# Read the rows that the chart with this answer id shows
kater answer 00000000000000000000000000000000
# Give the rows as JSON, for a script
kater answer 00000000000000000000000000000000 --json

Arguments

ArgumentDescription
<id>The answer’s id, as kater here lists it beside the file of a chart, table or panel. A run’s snapshot_id also works. Give one or more.

Options

It also takes the global options --json and --help.

The rows on screen

Kater reads the rows as you. It first does the page’s own check. You must still reach the query, and the query’s file and its workspace must not have changed. There is no build and no warehouse call. It needs kater login in your own terminal.

What it answers

The answer has the shape of a kater run result: result holds columns, fields and rows. It also gives these keys:

  • from: page when an open tab shows the rows, with the tab and what it shows. Else kept.
  • path, query and inputs: the query and the inputs that the rows come from.
  • ran_at: when the rows were read.
  • refreshing: the page reads the rows again.

When it cannot read the rows

An id that is not yours, or whose files changed, says why with a code. Run the query then.

  • answer_not_found: the answer is gone, or was never yours.
  • answer_not_yours: the answer was read for someone else.
  • answer_unreachable: you no longer reach the query.
  • answer_stale: the query’s file or its workspace changed since the rows were read.

Through the Kater MCP, the read_rows tool reads more rows than fit in one reply, page by page.

See also