Fixes Broken Links

Periodic Website Scan

The Hyperlint crawler will periodically scan your configured documentation website for broken internal links based on the frequency that you configure.

To minimize unnecessary web traffic, Hyperlint will always crawl your website according to the RFC 9111 web caching standards.

After the crawl is complete, it will move to the next phase: analyzing and fixing the broken links.

Broken Link Analysis and Fixes

Hyperlint will build an internal graph of your content and analyze the graph for broken internal links. The type of links that are currently supported are:

1. HTML fragment links on the same page: for example a link on getting-started.md with an href linking to getting-started.md#quickstart or just #quickstart
2. HTML fragment links between pages: for example a link on readme.md with an href linking to path/to/advanced-guide.md#kubernetes

Other types of internal links and external links are not supported yet, refer to the limitations section below.

Once Hyperlint has identified the broken links, it will use an AI to fix them based on a graph of the crawled content. If no broken links are able to be identified or fixed, it will wait until the next scheduled time before checking again.

If changes are detected, the Hyperlint will continue to the next step.

Pull request creation

If Hyperlint is confident in the changes that it proposes, it will create a pull request with the suggested changes. This will be a pull request to your repository and default branch.

The pull request will contain detailed context about each broken link and the reasoning behind choosing the fixed version of the link. If, after your review, you'd like to accept the changes, you can merge the pull request. If not, just close it out!