AmpliForgeBot is the automated web fetcher of AmpliForge GmbH (Eschborn, Germany). AmpliForge is a content platform for B2B marketing teams. If you see AmpliForgeBot in your server logs, one of our customers asked AmpliForge to read a public web page, most often the customer's own website.
When a team sets up its AmpliForge workspace, AmpliForge reads the public pages of the company's website to learn its offerings, tone of voice and terminology. Customers can also point AmpliForge to other public pages, for example for market research. AmpliForgeBot only requests publicly reachable pages.
AmpliForgeBot identifies itself with this User-Agent header: Mozilla/5.0 (compatible; AmpliForgeBot/1.0; +https://www.ampliforge.com/bot). If a website refuses this identification with a block or a rate limit, AmpliForgeBot may retry that page, and later pages of that website for 24 hours, with a browser-style identification. Requests from the same servers still come from AmpliForge's infrastructure.
AmpliForgeBot fetches pages one after another with a pause per host, not in bursts. It reads at most a few hundred pages per scan, and a site is not scanned again for several days. If your server answers with HTTP 429 or a Retry-After header, it slows down and waits. If your server keeps answering with blocks or challenges, it stops early instead of pushing harder. For third-party pages it honours the Disallow rules in robots.txt that apply to AmpliForgeBot or to all robots, and it skips pages that carry an AI or text-and-data-mining reservation.
Add these lines to your robots.txt: User-agent: AmpliForgeBot, then Disallow: /. This applies to pages we read as third-party sources, such as competitor research. A website owner who asks us to scan their own site is not affected by robots.txt rules. If you need something else, for example a lower request rate, write to us.
If AmpliForgeBot causes a problem on your site, or you would like it to visit less often or not at all, write to info@ampliforge.com with the affected domain and the time of the requests.