Skip to main content
Every command picks its output format from the terminal and the global flags. On a terminal you get a readable table or key-value list. Anywhere else, and whenever a format flag is set, you get machine readable output. When more than one flag is set the precedence is --toon, then --quiet, then --json, then the default.

Pretty output

Lists print as a table, single resources as a key-value list, and both end with breadcrumbs suggesting the command to run next:
A table shows at most seven columns. A fixed set of well known fields such as id, name, title, status, duration, and created_at comes first, in that order, then the remaining scalar fields in alphabetical order, and any URL fields last. Columns shrink to fit the terminal width and long values are cut with ..., except URLs, which wrap so they stay usable. In a key-value list, nested arrays and objects are summarized as (3 items) or (4 fields) unless they are short.

JSON envelope

data holds the resource or the array of resources. breadcrumbs is left out when a command has none, and pagination appears only on list commands. Commands that print a message rather than a resource, such as delete, login, and logout, emit it as {"message": "Recording deleted."} when piped or with --json.

Quiet output

--quiet drops the envelope and prints the payload alone. Commands that create a resource, such as folders create, recording-requests create, and recordings create-cta, print just the new id. Commands that print a message, such as delete, login, and logout, print success. Everything else, including the update commands, prints the raw JSON payload. That makes it the right choice inside a shell pipeline:

TOON output

--toon emits TOON, a token optimized encoding of the same data. It carries the same fields as the JSON envelope in noticeably fewer tokens, which matters when the output is going into an AI assistant’s context window.

Pagination

List commands take --page and --page-size, with a maximum page size of 100. The pagination block reports current_page_number, total_pages, total_records, and page_size, so a script can keep incrementing --page until it reaches total_pages.