Skip to main content

Using the catladder CLI

catladder is installed per project — invoke it as yarn catladder (never a globally installed copy). Every command can run non-interactively, which is the mode agents should use.

Invocation​

yarn catladder <command...> [args] [flags]

# examples
yarn catladder project doctor
yarn catladder project ci job-log
yarn catladder --help                 # command tree
yarn catladder project --help         # subcommands of a group

Passing inputs​

Commands declare typed inputs. Any input not provided on the command line is prompted for interactively — in a non-interactive context (agents!) provide them up front:

  • positional inputs: yarn catladder project k8s list-pods dev:api
  • flag per input: --some-input value (kebab-case of the input name), booleans as --flag / --no-flag
  • all at once as JSON: --inputs '{"envComponent": "dev:api"}'
  • -y / --yes skips confirmation prompts

A missing required input fails with a message naming the exact flag to pass. Environment/component selectors use the env:component form (e.g. dev:api, review-123:www).

Some commands are inherently interactive (e.g. project config-secrets opens $EDITOR) — ask the user to run those.

Command reference​

See references/commands.md for all commands with their descriptions and inputs (generated from the CLI itself, so it matches the installed version).

  • catladder-config — catladder.ts structure and regeneration
  • catladder-builds — the component build config
  • catladder-deploys — the component deploy config
  • catladder-secrets — env vars and secret management
  • catladder-releases — releasing (semantic-release / changesets)
  • catladder-pipelines — pipeline behavior and CI debugging