Rolit CLI
Rolit CLI
Run Rolit's agent from your terminal, in your own workspace, with approvals you control.
rolit is the Rolit command-line client. It signs you in, connects to a project, and runs Rolit's hosted agent against a folder on your machine. The agent can read and edit files and run commands in that folder, but only as far as your approval mode allows.
The CLI doesn't hold agent prompts, provider credentials, or other services' tokens. Those stay on Rolit's servers.
What you can do
- Chat or run a goal. Use
rolit chatfor an interactive session, orrolit run "<goal>"for a single task. - Work on marketing and content jobs. Research, content, creative, SEO, email, reports, and more run as
rolit <area> <job>commands. - Review risky actions. Anything that publishes, sends, or spends is held for your approval.
- Script it. Add
--jsonfor machine-readable output, and use exit codes in CI.
In this section
| Page | What it covers |
|---|---|
| Install | npm, Homebrew, shell script, PowerShell, updates |
| Quickstart | Sign in, pick a project, and run your first task |
| Approval modes | What the agent may do, and how pending actions work |
| Commands | The command groups at a glance |
| Slash commands | Every /command in rolit chat, plus keyboard shortcuts |
| Agent tools | How tools load and are approved, with pages for each family |
| Decision support | Classify, score, or filter every row of a file |
| Setup | Install managed components such as DuckDB |
| Local data | Run history, usage, and what is stored on your device |
| Automation | --json, environment variables, and exit codes |