FAQ

Common questions and troubleshooting

Extraction issues

Empty results

  • The page loads data dynamically after scroll. Add "scroll down to load all content" to the website description.
  • The page requires login. The agent may have landed on a login page. Set up a saved login.
  • Schema doesn't match the page. The agent can only extract what's on the page. If your schema fields don't correspond to visible content, results will be empty or null. Try running with a minimal schema first.
  • Slow-loading site. The agent waits for content, but extremely slow pages can time out. Add "wait for all content to load before extracting" to the description.

Wrong data extracted

  • Add descriptions to schema fields. Instead of "name": { "type": "string" }, use "name": { "type": "string", "description": "the full event name including series title" }. The agent uses these as hints.
  • Narrow the extraction scope. Add "only extract from the main results grid, ignore sidebar and header content" to the website description.
  • Watch the live view. Run a new extraction and observe where the agent is pulling data from. This quickly reveals mismatches between the page and your schema.

Run takes too long

  • Use a playbook. AI-driven runs explore the page and write scripts from scratch. Playbook runs skip exploration entirely — the same extraction typically runs 3–10x faster.
  • Reduce scope. If extracting from many pages, confirm the extraction works on a single page first.
  • Check browser resources. Local browser runs depend on your machine's available memory and CPU.

Browser issues

Page won't load

  • Check the URL. Confirm it's accessible in a normal browser from your network.
  • The site blocks automation. Use the local browser (Camoufox/Firefox) — it has significantly better anti-detection than Chromium-based cloud browsers.
  • Corporate firewall. If you're on a restricted network, some sites may not be reachable from your machine. Cloud browsers connect from external IPs and may work as an alternative.

Login not working

  • Session expired. Saved logins expire. Re-export your session from a fresh browser login and reimport it.
  • IP mismatch. Some sites bind session tokens to the IP that created them. Cloud browsers use different IPs and may trigger re-authentication. Switch to the local browser.
  • 2FA was required. Two-factor authentication prompts can't be automated. Complete the full login including 2FA in your browser, then export the resulting session.

Local browser server won't start

  • Port already in use. Run lsof -i :2026 to check. Stop the conflicting process or start the server on a different port with --port.
  • Not enough memory. The browser needs at least 2 GB of free RAM. Close other applications.

Live view not updating

  • Local browser uses polling-based screenshots updated every ~500ms. A small delay is expected and does not affect extraction speed.
  • Cloud browser uses a real-time iframe. If it appears stuck, refresh the page.
  • Run already finished. The live view is only active during running extractions.

Schemas and data

Can I reuse a schema across multiple websites?

Yes. Schemas are defined at the workspace level under Schemas in the sidebar. Attach the same schema to as many websites as you need — useful when you extract the same structure from multiple sources.

What happens if a required field is missing on the page?

The agent will set the field to null if it can't find a value. If the field is listed in required in your schema, the agent will retry the extraction with adjusted logic before accepting a null result.

Can the agent extract data from PDFs or images?

Not directly. Ev3ry is designed for HTML pages. If a site embeds data in a PDF or image, the agent can navigate to the PDF URL but cannot parse the content from it.

Account and workspace

Can I share access with my team?

Yes. Invite team members from Settings > Members. Members can view and manage websites, schemas, playbooks, and runs within the shared workspace.

How is usage billed?

  • Local browser: No browser cost (runs on your hardware). Ev3ry platform usage is based on your plan.
  • Cloud browsers: Billed directly by the provider — BrowserBase charges per-minute, Browserless pricing varies by plan.
  • See the Settings > Billing page for current plan details.

How do I delete my data?

  • Run results — delete from the run detail page
  • Saved logins — remove from the website settings
  • Entire website — delete from the website list (removes all associated playbooks, schemas, and runs)
  • Account — contact support for full account deletion
NextWelcome to Ev3ry