HyperNewsBot

HyperNewsBot is the reader behind Hyper News. It fetches news reports listed by The GDELT Project and reads them to count what happened. It identifies itself as HyperNewsBot in its user agent.

What it does with a page

• It reads the text to find figures (killed, injured, arrested and so on), the place and the time.
• Hyper News never shows the text, the headline, the author's words or the photos. It shows its own sentence for each figure and a link to the sentence in the report.
• The text is kept for at most 30 days so it can be read again after a fix, then deleted.

How it behaves

• It obeys robots.txt for HyperNewsBot and for *.
• It honours text-and-data-mining reservations: a tdm-reservation: 1 header, <meta name="tdm-reservation" content="1">, a /.well-known/tdmrep.json file (W3C TDMRep), or a robots meta tag with noai. A reserved page is not read at all.
• It asks each site for at most one page every few seconds and backs off when a site answers 429 or 503.
• It does not log in, does not get past paywalls and does not run a site's scripts.

To stop it

Add to robots.txt:

User-agent: HyperNewsBot
Disallow: /

or reserve text and data mining as above. To have links to your reports removed as well, use Corrections and removals and choose "I am a publisher".