On-demand headless browser sessions

Browserless spins up a headless Chrome instance on request instead of requiring a team to run and maintain its own browser infrastructure. A Beam agent opens a session when a trigger fires, such as a scheduled test or a request from another tool, runs the approved script against a page, and reads back the result: rendered content, a screenshot, or a status code. Sessions that time out, hit a page error, or return an unexpected status are closed and routed to a person to check the target site or script, rather than retried indefinitely.

Scripted page interaction and screenshots

Browserless can execute a script against a live page, click through steps, fill fields, and capture a screenshot or PDF of the result. A Beam agent runs the approved script on a trigger, reads the captured output, and checks it against the expected result, such as confirming a page loaded correctly or a form step completed. Where output does not match, for instance a login page appears instead of the expected content, the agent stops the run and sends the case to a person along with the captured screenshot so they can see exactly what happened.

Concurrent session and usage tracking

Because Browserless charges and limits usage by concurrent sessions and browser time, a Beam agent reading account or session data can track how many sessions are running and how close usage is to any limit the account permissions expose. On a schedule, or when a job is queued, it checks capacity before starting new sessions and holds jobs back if the account is near its limit. When usage approaches a threshold the customer has set, the agent notifies a person rather than letting jobs fail silently.