The tool parses your Markdown content to extract all hyperlinks — including standard [text](url) links, image references ![alt](url), bare URLs, and anchor links like [text](#section). It then attempts to verify each external URL using HTTP requests. For anchor links, it cross-references them against headings found in your document. Results are categorized as valid, broken, or needing manual review (e.g., when CORS policies prevent automated verification).

Many websites enforce CORS (Cross-Origin Resource Sharing) policies that block browser-based requests from external origins. When this happens, our tool cannot determine the actual HTTP status code. These links are flagged as "Needs Review" so you can manually verify them. Additionally, relative paths (like ./docs/page.md) and special protocols (mailto:, tel:) cannot be checked automatically and are marked accordingly.

The checker handles: External HTTP/HTTPS links (verified via fetch requests), Internal anchor links (e.g., #introduction — validated against document headings), Image links (![alt](url)), Bare URLs (plain https://...), Relative paths (./file.md, ../folder/doc.md), and Special protocol links (mailto:, tel:, ftp://). Each type receives appropriate handling.

The tool scans your Markdown for all heading lines (lines starting with #) and generates slugs following GitHub Flavored Markdown conventions: text is lowercased, special characters are removed, spaces become hyphens, and consecutive hyphens are collapsed. Anchor links like [Go to setup](#initial-setup) are then matched against these generated slugs. A match means the anchor is valid; otherwise, it's flagged as broken.

Yes! The tool processes links in batches with concurrency control (up to 5 simultaneous requests) to avoid overwhelming your browser or triggering rate limits. However, very large files with hundreds of external links may take a while to fully validate. A progress bar keeps you informed throughout the process. For files with 100+ links, expect the check to take 30–90 seconds depending on network conditions.

A broken link definitively returned an HTTP error (4xx client error like 404 Not Found, or 5xx server error) or failed to resolve entirely (DNS failure, timeout, network error). A link that needs review is one where the tool received a response but couldn't confirm its status — most commonly due to CORS restrictions. These links may still be perfectly valid; they simply require manual clicking to verify.

For external broken links: verify the URL is still correct, check if the target page has moved (look for redirects), or find an archived version. For broken anchor links: ensure the heading text in your document exactly matches the anchor slug. For relative path links: confirm the target file exists at the specified location. Use the "Copy Broken" button to export all broken URLs and work through them systematically.

When a URL responds with a 3xx redirect and the server permits CORS, the tool's fetch requests follow redirects automatically and report the final destination's status code. However, if CORS blocks the initial request, redirect information cannot be retrieved. In such cases, the link is marked as "Needs Review." Modern browsers' fetch API follows up to 20 redirects by default.