Findkit’s new link report lists the internal links on your site that lead nowhere, and it does it
without sending a single request to your server.
Broken links accumulate quietly
Pages get renamed. Sections get restructured. Old campaign pages get deleted. Every one of those
changes can leave a link somewhere else on the site pointing at an address that no longer works.
The reason nobody notices is that nothing appears to be wrong. The page containing the broken
link loads perfectly. It looks finished. The failure only shows up for the visitor who clicks,
and they rarely tell you. So the count creeps up over months, and the first real signal is a
customer email or a drop in the analytics for a page you thought was still reachable.
Checking by hand does not scale past a few dozen pages, and checking by tool has usually meant
pointing a link checker at your site and asking it to fetch every page all over again.
Findkit already knows
This is the part worth understanding, because it is what makes the report different.
Findkit crawls your site to build your search index. That crawl already visits your pages and
already records the outcome of every address it fetches. The link report is assembled from data
Findkit collected on that pass.
So generating the report sends no requests to your site at all. There is no second crawl, no
traffic spike in your logs, no load on your server, and nothing to schedule for a quiet hour.
The work was already done. You are reading the results of it.
What the report shows
Findings are grouped by the page the link was found on, which matches how the work actually gets
done: open a page, fix its links, move to the next one.
Links are sorted into four groups:
- Not found. The target returns 404 or 410. The page it pointed at is gone.
- Redirects. The link works, but only via a detour. The report names the address the redirect
leads to, so you can point the link straight at the destination instead. - Server errors. The target returned a 5xx or could not be loaded.
- Forbidden. The target returned 403. Often this is deliberate, such as a page behind a
login, so these are worth reviewing rather than fixing blindly.
Redirects deserve the extra attention. They are easy to dismiss as harmless, since the visitor
still arrives, but every one of them is a wasted round trip and a piece of history nobody has
tidied up. Because the report gives you the final destination, fixing them is a copy and paste
rather than an investigation.
The report tells you when the crawl behind it finished. The figures describe your site as of the
last completed full crawl, so you can see at a glance whether you are looking at this morning’s
picture or last week’s. It also flags when that crawl was incomplete, so a partial report cannot
be mistaken for a clean site.
Two ways to read it
In the Findkit Hub. Open your project, go to Inspect, then Links. Nothing to install and
nothing to configure beyond the settings below.
Over the REST API, if you would rather have the data in your own dashboard or reporting:
GET https://api.findkit.com/v1/projects/{projectId}/link-report
The endpoint is API-key authenticated and takes optional target and limit parameters. Full
reference in the REST API documentation.
Switching it on
The link report is built from a record of every link on every page, and collecting that record is
a Findkit Pro feature. On Pro, it takes a few settings on your target in findkit.toml:
track_links = "all"
walk_links = true
use_sitemap = true
See the documentation: track_links and walk_links
Add them, deploy, and run a full crawl. The report is assembled from crawl data, so the first
findings appear once your site has been crawled with the settings active. This is a small job for
whoever maintains your site.
Without track_links, Findkit tracks only PDF links and the report will look almost empty. That
empty report means the setting is off, not that your site is clean, which is a distinction worth
knowing before you celebrate.
What it does not do
It covers links whose target Findkit fetched, which in practice means links to your own pages. A
link out to somebody else’s site is not in this report, and a clean report here does not mean
every external link on your site still works.
It also finds and reports. It does not edit anything. Fixing a link is still a person opening a
page and changing it.
Getting started
If you are already on Pro, add the three settings, run a crawl, and open Inspect → Links.
If you are on another plan and this is the thing you have been missing, Pro subscription is where it lives.