Skip to content

Human in the Loop

If a headless scrape hits a CAPTCHA or block page, automatically spawn a visible browser for human intervention, then resume extraction.

name = "Hybrid Recovery"
url = "https://example.com/protected-page"
selector = ".data-item"
fields = [
"title:.title",
"value:.value"
]
interval = 600
  • store_* persists recovery state across runs
  • Headless browser runs normally until a block is detected
  • Visual browser spawns for human intervention
  • before_fetch checks if the puzzle was solved before continuing
  • Timeout resets the cycle if no one intervenes