Report a bug
Something returned the wrong result, or would not run at all? Tell us which tool, what you gave it, and what you expected.
Last updated
What helps most
A small example that reproduces it. Two lines of text, a fragment of markup, a three-entry sitemap. If you can make the problem happen with example.com addresses, that is ideal.
What you expected. "It found three links but there are four" tells us far more than "it doesn't work". So does "it split this address in half".
Which tool and which mode. The website tool has single-page and crawl modes; the YouTube tool has three. They share almost no code.
Your browser and operating system, if something looked broken rather than wrong.
The reference code, if you saw an error page. It lets us find the matching log entry.
Known limits that are not bugs
Before reporting, it is worth checking whether you have hit a documented limit rather than a defect:
- A crawl cannot find pages that nothing links to, or pages that only appear after JavaScript runs.
- OCR cannot recover a link hidden behind the words "click here" — the address is not in the image.
- A scanned PDF has no text layer and no link annotations to read.
- The YouTube playlist mode needs an API key to be configured for this site.
- Bare domains in the text tool are off by default, because they are ambiguous.
Each tool page has its own "What this tool cannot do" section.
What happens next
Your report is stored so we can work on it, and forwarded by email if this site has a mail transport configured — the confirmation tells you which. There is no ticket number and no automated reply; a person reads it.
Report a bug
Tell us which tool misbehaved, what you gave it, and what you expected.