Skip to content
Skryp

A live browser for pages that need clicks, or a sign-in

For filters, forms and multi-step pages, open a live browser session from a chosen country. Each snapshot lists the page's interactive elements with refs your agent can click or type into, then read the result as Markdown or a screenshot. For your own accounts, connect a login once in the dashboard: Skryp keeps only that site's sign-in, sealed for your workspace, and your agent can read pages and run saved workflows as that account.

  • Numbered element refs for reliable actions.
  • Screenshots and Markdown of the current state.
  • Saved logins: sign in once through a live browser; several accounts per workspace, each callable by name.
  • Workflows: repeat clicks, filters, pagination and downloads with inputs, checkpoints, limits and a schedule.
  • A signed-out page is reported as needs_login, never returned as data; paused runs continue after you reconnect.
  • Sessions close after 10 idle minutes.

Limits

  • Consequential actions (purchases, submissions) are yours to approve; Skryp does not decide them.
  • Skryp never stores what you type when signing in; a site that asks for a password or code again needs you to reconnect.
  • Use logins only on sites whose terms allow automated access to your account.

Price

Opening and each navigation cost one page: 1 credit on our network, 10 through a residential proxy. Reading and clicking are free. Full rate card

Open a session · REST
curl -X POST https://api.skryp.dev/v1/browser \
  -H "Authorization: Bearer $SKRYP_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "url": "https://shop.example/search",
  "country": "us"
}'
Open a session · MCP tool call
{
  "tool": "browser",
  "arguments": {
    "url": "https://shop.example/search",
    "country": "us"
  }
}

Related