Skip to content
TidyKB
Health scan

Every broken link and stale article, listed.

A read-only scan of every category, folder, solution article and translation in your Freshdesk knowledge base, with a health score and issues ranked by how many customers see them.

Run free audit
Sample TidyKB issue list for a Freshdesk knowledge base: “Reset your password” links to a deleted article, “Set up SSO” mentions 2023 pricing, the Billing FAQ has an image that returns 404, the German “Export your data” is three versions behind, two titles are near-duplicates and “Connect Slack” has a dead external link.

The problem

Freshdesk stores your articles. Nobody checks them.

You archive or merge an article and the links to it break. A rename leaves dozens of articles with the old name. Freshdesk shows views and votes, but it has no broken-link check, no duplicate detection and no list of stale articles. So the knowledge base gets reviewed in a spreadsheet once a quarter, if at all.

How it works

How health scan works

  1. 1

    Connect read-only

    Create a dedicated agent with a knowledge-base-only role, paste your domain and its API key, and test the connection.

  2. 2

    TidyKB reads everything

    Every category, folder, article and translation is fetched within your Freshdesk rate limit. Links and images are checked in batches.

  3. 3

    Work the list

    Filter by issue, folder, language and severity. Fix, ignore or assign, and watch the score move.

Why it matters

What you get

Links Freshdesk-aware, not just 404s

Catches links to your own deleted, draft or unpublished articles, which a generic crawler sees as fine.

Stale content, flagged

Articles not updated in N months, old years and old product names in the text, empty folders and drafts never published.

Read-only by default

The scan only reads Solutions endpoints. It never touches tickets, contacts or settings.

Before and after

Today in Freshdesk, and with TidyKB

Today in Freshdesk

  • Today: Click through every article to find dead links
  • Today: Sort by “modified” and guess what’s stale
  • Today: Hear about duplicates from a confused customer

With TidyKB

  • Every broken link and dead internal link, listed
  • Stale articles flagged by age and by old years or names
  • Near-duplicate titles shown in pairs
GuideHow to find broken links in FreshdeskFive ways links break in a Freshdesk knowledge base, a spreadsheet method for small help centers, and how to prevent it.

FAQ

Questions, answered

Which checks run?

Broken external links and images, links to deleted, draft or unpublished articles, articles not updated in a set number of months, old years and product names, near-duplicate titles, empty folders, drafts never published, and translations older than their primary article.

How often does it scan?

Weekly on Starter and Team, daily on Pro. The daily scan on Pro re-reads only what changed since the last one; the weekly scan always reads everything.

Why a dedicated agent?

A Freshdesk API key carries the permissions of the agent it belongs to. A dedicated agent with a role limited to the knowledge base keeps TidyKB away from everything else, and you can deactivate it any time.

How is the health score calculated?

Each issue costs points by severity: broken links and links to deleted articles cost most, stale content less, near-duplicates least. One article can only lose so many points however many problems it has, so a single bad article can’t sink the score, and the total is measured against the size of your help center — so a 100-article and a 2,000-article knowledge base compare fairly. A scan that couldn’t read everything says so instead of scoring what it never saw.

See your help center’s score.

Paste a URL. No signup, no API key, no call.

Run free audit