README

A Kirby CMS plugin that crawls a website, discovers all links, and checks them for broken HTTP responses. Results are displayed in a Panel section or via the CLI.

The crawler follows all same-domain links to discover pages, then verifies every link it finds: both internal and external. It supports seeding from sitemap.xml files.

Requirements

• PHP 8.3+
• Kirby CMS 5.3+
• Kirby CLI (optional, for the kirby linkcheck command)

Installation

composer require moinframe/kirby-linkcheck

Usage

There are two ways to use this plugin:

1. Panel section: add a section to any blueprint and check links directly from the Kirby Panel
2. CLI command: run kirby linkcheck from the terminal

Panel Section

Add the section to any blueprint:

sections:
  links:
    label: Check for broken links
    type: linkcheck
    defaultUrl: https://example.com # optional, defaults to site URL
    defaultSitemap: https://example.com/sitemap.xml # optional
    editable: false # optional, disables URL fields (default: true)

How It Works

1. Enter a URL and click Check Links
2. The crawler runs and returns results directly
3. A results table shows all non-200 links with status, source page, and linked URL

CLI Command

Check your site for broken links from the terminal:

# Check a URL
kirby linkcheck https://example.com

# Check a URL with a sitemap
kirby linkcheck https://example.com https://example.com/sitemap.xml

# Provide only a sitemap (base URL is derived from the sitemap host)
kirby linkcheck https://example.com/sitemap.xml

The command crawls the site, then prints a summary and lists all broken links:

Checking links on: https://example.com

  [404] https://example.com/about → https://example.com/missing-page
  [0] https://example.com/contact → https://dead-link.example.com

Done. 42 links checked, 2 broken.

Status code 0 means the connection failed entirely (DNS error, timeout, etc.).

URL Handling

• Relative URLs are resolved against the page they were found on
• Fragment identifiers (#section) are stripped
• mailto:, javascript:, tel:, and data: schemes are skipped
• . and .. path segments are resolved
• Protocol-relative URLs (//cdn.example.com) inherit the base scheme
• Duplicate URLs are deduplicated per page

Same-Domain Detection

A link is considered same-domain if its host matches the start URL's host exactly. Subdomains are treated as external (e.g., if the start URL is example.com, links to www.example.com are external).

PHP Configuration for Large Sites

For sites with many pages you may need to adjust your PHP configuration:

; php.ini or pool-specific config
max_execution_time = 300   ; Allow up to 5 minutes
memory_limit = 512M        ; Increase if errors occur

The CLI command is not affected by max_execution_time (PHP CLI has no time limit by default).

License

MIT