Files & Clipboard | agent-browser

agent-browser can upload files, capture downloads, read local files through file:// URLs, write PDFs and screenshots, and interact with the browser clipboard.

Upload files

agent-browser snapshot -i
agent-browser upload @e4 ./invoice.pdf
agent-browser upload @e4 ./front.png ./back.png

The selector must resolve to a file input. Multiple file paths are accepted for multi-file inputs.

Downloads

agent-browser download @e5 ./report.csv
agent-browser wait --download ./archive.zip --timeout 30000

Use download when a specific element triggers the download. Use wait --download when another action starts the download and you need to wait for completion.

Set a default download directory for browser-initiated downloads:

agent-browser --download-path ./downloads open https://app.example.com

Without --download-path, downloads go to a temporary directory that is cleaned up when the browser closes.

Screenshots and PDFs

agent-browser screenshot ./page.png
agent-browser screenshot --full ./page-full.png
agent-browser screenshot --if-changed
agent-browser screenshot --threshold 0.01
agent-browser screenshot --screenshot-format jpeg --screenshot-quality 80 ./page.jpg
agent-browser pdf ./page.pdf

Prefer --if-changed for repeated captures: skipping unchanged images is the most token-efficient option. History is per tab and capture scope. The first capture returns a path; later unchanged captures omit it. JSON includes changed, revision, and pixelChangeRatio. --threshold <0-1> implies --if-changed and treats changes at or below that ratio as unchanged.

Screenshot defaults can also be configured with:

Setting	Description
`--screenshot-dir` / `AGENT_BROWSER_SCREENSHOT_DIR`	Default screenshot output directory
`--screenshot-format` / `AGENT_BROWSER_SCREENSHOT_FORMAT`	`png` or `jpeg`
`--screenshot-quality` / `AGENT_BROWSER_SCREENSHOT_QUALITY`	JPEG quality from 0 to 100

Local files

agent-browser --allow-file-access open file:///Users/me/report.pdf
agent-browser --allow-file-access open file:///path/to/page.html
agent-browser screenshot ./local-file.png

--allow-file-access is Chromium-only. It allows file:// pages to load and access other local files through browser APIs such as fetch and XHR.

Clipboard

agent-browser clipboard read
agent-browser clipboard write "Hello, world"
agent-browser clipboard copy
agent-browser clipboard paste

copy and paste simulate the platform keyboard shortcuts for the current selection or focused element. Use write when you want to set clipboard text directly.