Websites

Managing data sources in Ev3ry

A website is the top-level entity in Ev3ry. It represents a data source — a URL the agent navigates to collect structured data.

What a website contains

FieldDescription
URLThe starting page for the agent
NameA label for your own reference
DescriptionWhat data you want (guides the AI's exploration)
Saved loginOptional session data for authenticated pages
SchemasThe output structures attached to this website
PlaybooksSaved extraction workflows ready for replay
RunsThe full history of extraction attempts

Adding a website

From the Websites page or the dashboard, click Add Website. You need at minimum a URL and a name. The description is optional but strongly recommended — the agent uses it to understand the goal of each extraction.

Good description: "Extract all upcoming fixtures from the Premier League schedule, including home team, away team, date, time, and venue."

Less useful: "Get the data from this page."

Website status

Each website shows a status based on its most recent run:

  • Ready — last run succeeded; playbooks are up to date
  • Learning — an AI run is in progress (no playbook yet, or playbook is being updated)
  • Error — last run failed; playbook may be stale

Settings

Under a website's settings you can configure:

  • Default schema — used when creating new runs
  • Default browser — local or a specific cloud provider
  • Saved login — session data for authenticated access

Organizing websites

The sidebar groups your websites by status. Use the search and filter controls on the Websites page to find specific sources across large workspaces.