Identify why a website import failed, grant scoped access, test it, and finish learning.

What you’ll accomplish

Start with the explanation on your website source. Follow the matching check below, then return to Chat4U to verify the import.

Finish with: Readable website content, or a usable file/text alternative

Start with your affected source

Collect the details of this attemptUse the exact address and time shown by Chat4U.
Open Project → Knowledge, find the affected website source, then select Resolve website access or Website access settings. Read the explanation, the affected-page count and Request details. Note the affected address, observed time and timezone, and provider request ID if one is shown. A Chat4U diagnostic reference identifies our attempt; it may differ from your provider's request ID.
If the latest refresh failed, check the source’s last successful synchronization time. Your previously imported pages remain available while you recover. A partial import needs attention only for the affected pages.
Choose who will change website settingsYou can continue even if someone else manages the site.
If you manage the website, follow the matching section below. If your developer, agency or host manages it, open Request details → Copy instructions for my administrator and share the safe report with them. It includes a relevant guide; it does not send a message automatically. Keep tokens, passwords and session cookies out of the report.
You can use a file or trusted text while waiting. You do not need to delete and recreate the website source.

Choose the matching check

What Chat4U says	Start here
Security verification, firewall block or unexplained refusal	Website security refused access
This page requires authorized access	Login required
Could not connect or verify the secure connection	Connection and certificate errors
Website timed out or returned an error	Timeouts and website errors
Page not found or no readable content	Missing or unreadable pages
Chat4U could not complete the attempt	Chat4U errors
The cause has not been confirmed	Unknown cause

Website security refused access

Use the detected provider's guide: Cloudflare, AWS WAF or Vercel. A provider name identifies where to investigate; it does not prove which security rule caused the refusal.
If no provider is identified, ask your administrator to find the affected request in hosting/security logs using the address and time. Ask which service or rule refused it before creating an exception.
Use only a method that the provider and this Chat4U source support. A Allow the Chat4U crawler identity matches the public Chat4UBot/ User-Agent prefix. A Website access token matches a private header value configured on both sides. Importer IP addresses are usable only when Chat4U displays a current verified list.
After the administrator saves the matching exception, verify and finish. A plain HTTP 403 alone does not establish bot protection.

For an IP-based exception, open your source's Website access settings, select Access method → Allow verified importer IP addresses, then Copy IP addresses. Share that list with your administrator; use another method if IP access is unavailable.

Open Chat4U

Message to your website administrator

Please help restore the Chat4U content import for the address in the attached report.
Find the request at the report's time and timezone; use its provider ID if present.
Tell me which service/rule refused access and configure only the required exception.
The User-Agent prefix is Chat4UBot/; it is public and can be copied by others.
If we use a private header token, arrange it privately and configure the same value
in the provider and Chat4U. Do not add the token to this report or the website URL.
After the change, I will run Test access and retry import from the same source.

Attach the sanitized report from Chat4U. An absent request in sampled logs is not proof that the provider allowed it; check the correct time, hostname and available logging.

Open the affected URL in a private browser window without signing in. If it asks you to log in, your normal browser session may be why the page works for you.

For supported Vercel automation access, follow the Vercel procedure.
A custom header token works only when your website administrator configures the website to accept it. Creating an arbitrary token in Chat4U does not make a login-protected page readable.
For an ordinary member/account login, use an intended public page or export the authorized content into a supported file/text source. This flow does not sign in with your website password or import your browser session.

You can proceed when: a supported automation method permits the intended content, or your alternative content source is processed and inspected.

Connection and certificate errors

Check the exact website addressConfirm the hostname and intended content path.
Open the affected URL in a private browser window. Correct a typo or retired address through the source row's Actions → Edit configuration control. If your public site uses a different hostname after a redirect, use the intended canonical address. A changed secure origin may require new access settings.
Ask the host for the matching checkA browser success alone does not prove the importer can connect.
For a connection failure, share the address/time and ask the host to check public DNS and incoming connections from the importer. For a certificate failure, ask them to check hostname coverage, expiry and the complete certificate chain on the exact hostname. Keep secure-connection verification enabled. Do not proceed through a browser security warning as a workaround.

You can proceed when: the address is correct and the hosting issue is fixed. Run the Chat4U access check again to verify its request path.

Timeouts and website errors

Open the affected page and check whether it returns the intended content. If it is unavailable, slow or shows a server error, give its URL and observed time to your host. Ask them to investigate that request, rather than add a bot exception from a timeout alone. When they confirm recovery, select Test access and retry import. If their browser works but Chat4U still fails, give them the updated request details. You can import trusted file/text content in the meantime.

Missing or unreadable pages

For Page not found, open the exact affected URL, find the replacement page if it moved, and update the source row through Actions → Edit configuration. For sitemap imports, ask the website administrator to update stale entries in the sitemap too. Keep other usable imported pages.

For No readable content, check whether the URL contains actual article/product/help text, rather than a loading screen, embedded image, download or login page. If the content is in a document, upload that document where supported. Otherwise paste the authoritative text, or give the administrator a representative affected page to investigate its rendering. Then inspect what the next import retained.

Chat4U errors

Use Re-sync on the affected source. If it fails again, retain the new diagnostic reference and use the product's existing help channel if available. Your website may be working normally; a Chat4U service error does not call for a website firewall change. File/text content is an alternative when available, and it also needs to finish processing.

Unknown cause

Check the exact address first. Share the safe report with your administrator to correlate the request with hosting/security logs. If neither side has enough evidence, keep the cause unconfirmed and use a direct-content alternative. Do not describe the failure as Cloudflare, a login wall or a firewall rule without matching evidence.

Verify and finish

Test from the same sourceReadable content starts the import; it does not finish processing.
Return to the affected source. Select Save settings if you changed the access method or token. Confirm you are authorized, then select Test access and retry import (or Test access and start import for a source that has never started). Access test queued means the saved check is waiting for the browser worker. Checking one page with your saved access settings… means it is running. Access check passed. Your import is queued or already running means the sample was readable and an import was accepted. You do not need a second start action.
Respond to the resultFollow the new explanation if the check is still blocked.
If access remains blocked, compare the latest URL/time/provider event with the rule you changed. If the check is inconclusive, confirm it reached the intended text rather than a challenge or login page. If Chat4U could not finish the check, use its stated retry path. After changing access settings, check again; an older successful test no longer confirms the current settings.
Inspect the completed importConfirm the pages your assistant actually needs.
Wait for the source's completed processing state. Open Pages and inspect representative content. Review any remaining affected pages separately; one successful sample does not cover an entire crawl or sitemap. Confirm assignment in Assistant → Knowledge, then ask representative questions. Retained old content stays usable during a failed refresh.

Import a file or paste text

In the same project's Knowledge, select Add source. Choose File, select a supported PDF, DOC, DOCX, TXT or MD and select Upload when available. Alternatively choose Text, enter a source name and authorized content, then select Add Text. Wait for processing to complete, inspect it, and assign it in Assistant → Knowledge. During guided setup, stay in the same project's Knowledge step. Follow the existing availability explanation if a file option is unavailable. File/text sources need their own updates when your website changes.

Last updated on

Get your website import working again