CLI reference#

This section is generated from the application itself, so it matches sktime-cli --help for the version you’re reading. Each command group has its own page, listed in the sidebar and at the bottom of this one.

Global options are accepted before any subcommand. --format and --json are also accepted on every leaf command, where they override the global value and so are not repeated in each command’s option list.

`sktime-cli`#

Command-line interface for sktime, designed for AI agents and humans.

Discover estimators (registry), fetch datasets, inspect and convert time series files (data), and run one-shot fit/predict/evaluate workflows. All commands support –format json for machine-readable output.

sktime-cli [OPTIONS] COMMAND [ARGS]...

Options

--format auto|human|agent|json|quiet: Output format: auto|human|agent|json|quiet. (default: auto, environment variable: SKTIME_CLI_FORMAT)
--json: Shorthand for –format json.
--cache-dir PATH: Workspace/cache directory (default: $SKTIME_CLI_HOME or XDG cache).
--no-cache: Bypass the registry disk cache.
--version: Print sktime-cli version and exit.
--install-completion: Install completion for the current shell.
--show-completion: Show completion for the current shell, to copy it or customize the installation.

Subcommands

cache: Manage the sktime-cli cache dir.
catalogues: Browse benchmark catalogues.
check: Run sktime’s estimator contract checks against an object spec.
data: Inspect, convert, split data files.
datasets: List and fetch datasets.
doctor: Check that sktime-cli is healthy: imports, cache dir, optional deps.
env: Report system info and versions of sktime’s dependencies.
metrics: List metrics and score results.
model: Inspect saved model artifacts.
registry: Discover sktime objects.
run: One-shot fit/predict/transform/detect/evaluate.
version: Show sktime-cli, sktime, and python versions.