How do I check outbound links on multiple pages at once?
Paste one URL per line into the urls_to_check list before the run - it ships pre-filled with three example URLs you should replace. The launcher builds the run queue from that list and processes it with up to five concurrent workers, so a few dozen URLs finish in minutes. Every URL produces one row in Outbound_Links and one row in Link_Census, de-duplicated by page URL.
What is the difference between internal and external links in an audit?
The census script resolves each href against the page origin. A link whose hostname equals the page hostname - including the www variant on either side - counts as internal; anything else resolving over http or https counts as external. Fragment, mailto, tel, javascript, data, blob and empty targets are tallied in their own columns and never counted as external.
How are mailto and tel links counted?
They are tallied, never copied. Mailto_Links and Tel_Links columns record how many such anchors exist on the page, and the Link_Census aggregate keeps matching counts. The external-samples field stores only http and https destinations, so no email address or phone number is written into any output file.
What does the max outbound links threshold do?
max_outbound_links (default 100) is the routing threshold applied after extraction. Pages whose Total_Links is strictly greater than the threshold land in the High_Link_Pages dataset; pages at or below it land in Good_Link_Partners. Lower the value to filter harder - for example, set it to 50 when you only want genuinely link-dense templates flagged.
How do I find pages with too many links?
Open the High_Link_Pages export and sort by Total_Links descending. Every row there exceeded your threshold, and the Absolute_HTTP and Root_Relative columns show whether the density comes from off-site links or from internal navigation markup. Empty High_Link_Pages with a populated Good_Link_Partners means no page crossed the threshold - a result, not a failure.
Does the checker follow links to other pages?
No. It audits exactly the URLs you supply and never crawls onward. It counts anchors on each page but does not visit their destinations, so it cannot tell you whether an outbound target is live, redirected or 404. It answers structure and page-health questions; destination status checking is a different job.
How do I export the link audit to CSV?
In standalone mode each dataset is saved as a local CSV file named after the dataset: Outbound_Links, Link_Census, High_Link_Pages and Good_Link_Partners. CSV, JSON and XLSX exports are all available from the launcher, and the files open directly in Excel or Sheets with no conversion step.
What navigation details does it record per URL?
Per URL the Outbound_Links row carries HTTP_Status, Final_URL, Redirect_Count, Response_Time_ms, Content_Type, Server, Protocol, Navigation_OK and Error - the same metadata a browser network panel shows, captured at run time. These values come from the engine's own navigation of your URL, not from any outbound link on the page.
Can it audit pages behind a login?
Not out of the box. Each run starts a fresh browser session with no stored credentials or cookies, so it sees what a logged-out visitor sees. If a URL redirects to a sign-in page, the audit records that final URL and its link structure - which usually looks obviously wrong in the census - rather than silently passing.
How are redirect chains reported?
Redirect_Count records how many hops occurred between the URL you requested and Final_URL, the address the browser settled on. A count of zero means the URL resolved directly; one usually means a normalisation hop such as adding www or upgrading to https; two or more suggests a chain worth flattening. Compare Page_URL against Final_URL to see exactly where the request ended.
Is counting links on a page allowed?
For pages you own or have permission to test, yes - this is a standard technical audit. For third-party pages, review their terms and robots.txt first, keep your URL lists proportionate, and remember five concurrent workers already load pages faster than a human browses. The bot collects only structural counts and public response metadata, no page content.
Does the bot need an internet connection to run?
Yes on first run: the launcher fetches its configuration by id from the backend at launch, so an internet connection is required at least the first time it starts.
Can I rename the downloaded file?
No. The downloaded file must keep its exact name - the filename carries the per-bot configuration id, and renaming it breaks the configuration fetch at launch.
