Quickstart

The fastest way to start rendering from your terminal. For the full reference, see the rest of the CLI guide.

Install and authenticate

npm install -g @urlbox/cli
urlbox login

urlbox login opens your browser. Approve, and the CLI stores your session and your project's render credential. In CI, set URLBOX_API_SECRET instead. Other install and auth options are in Install and Authentication.

Here it is in action:

Capture a full-page screenshot

--full-page captures the whole scrolling page; --output saves it to your current folder.

urlbox screenshot https://urlbox.com --full-page --output page.png

Save a page as a PDF

pdf is shorthand for render --format pdf --full-page.

urlbox pdf https://stripe.com/pricing --output pricing.pdf

Validate before spending credits

--dry-run checks your request and returns immediately without rendering: no API call, no credits used. Handy while you're getting a command right.

$ urlbox screenshot https://github.com --full-page --dry-run
✓ Dry run: payload validated, no API call made

Print the API call behind a command

--curl prints the exact API request a command would make, so you can drop it into a script or hand it to a teammate (or an agent). No render happens.

urlbox screenshot https://example.com --curl

Install the agent skill

Install the skill once and your AI coding assistant can drive the CLI for you in plain English.

urlbox skill install --target claude-code

Use any API option with --json

The CLI gives the common options their own flags. For anything else, including options without a flag, pass --json, and it's sent to the API verbatim.

urlbox render --json '{"url":"https://stripe.com","thumb_width":600,"format":"png"}' --output thumb.png

See Any option with --json for more, and the render options reference for the full list.

Where to next

• Rendering — every format, preset and flag.
• Screenshots from a CSV — render a whole list of URLs at once.
• Using the CLI from an AI agent — the full agent workflow.