Skip to main content

Kobana CLI

Kobana in your terminal.

One command for the APIs of every Kobana product. Built for people, scripts and AI agents: JSON output, schema at hand, and a dry run before you write.

Let your AI set up the environment

It installs the docs MCP, the CLI and the skills, and guides the login — asking for your confirmation along the way.

Agent with a terminal (recommended)

Installing the CLI and logging in happens on your machine. Run in your terminal:

Claude Code
$ claude "Follow the instructions at https://cli.kobana.com.br/en/getting-started/ai-setup.md to set up my Kobana CLI environment."
Codex
$ codex "Follow the instructions at https://cli.kobana.com.br/en/getting-started/ai-setup.md to set up my Kobana CLI environment."
Gemini CLI
$ gemini "Follow the instructions at https://cli.kobana.com.br/en/getting-started/ai-setup.md to set up my Kobana CLI environment."

Web chat (guidance only)

They do not run commands: they explain what to do and you run it.

Gemini does not take the prompt from the URL: the click copies the text and opens the site — just paste it.

Install in a minute

A single binary, no dependencies. Then log in to the product you will use.

$ curl -fsSL https://raw.githubusercontent.com/universokobana/kobana-cli/main/scripts/install.sh | bash
$ kobana billing auth login

Homebrew, Nix, ready-made binaries and source. All ways to install →

Made for automation

Always in sync with the API

Commands are built from the OpenAPI specs embedded in the binary. New API, new command — no hand-written CLI release to wait for.

Output for machines

JSON by default, NDJSON when paginating, table and CSV when you want them. Errors have their own exit code.

Schema in the terminal

kobana schema shows the parameters, body and responses of any endpoint without opening the docs.

Try before you write

--dry-run shows the request without sending it, and the idempotency key keeps a repeated change from duplicating.

Portuguese, English and Spanish

Help, messages and endpoint descriptions in your system language.

For AI agents

Inputs validated against injection, --fields to save context, and ready-made skills per product.

Built for AI agents too

Ready-made skills per product, a Gemini CLI extension, schema in the terminal and --dry-run on everything that writes. An agent learns what to do without reading the whole API.

What is in this documentation

SectionFor whomWhat you will find
DocumentationAnyone using kobana in the terminal, in scripts or with agents.Installation, authentication, syntax, output, pagination, schema, helpers and reference.
ManualAnyone looking for a command, a flag or a parameter.Every command of the seven products, with HTTP method, parameters, body and responses.

Ready for your first command?

Install it, log in to a product and list what is already in your account.

Get started