The PrivlekaiBot robot

If a visitor with this name shows up in your website log, it is our robot:

PrivlekaiBot/1.0 (+https://privlek.ai/bot)

It belongs to Privlek.ai, an assistant for small businesses in Kazakhstan. Below: why it visits, how it behaves and how to stop it.

Why it visits

A user of the service can request an express check of a website: their own, or someone else's, for example a prospective client's. The robot opens the public pages and records for the report:

• the response code and some headers (content type, caching, X-Robots-Tag and similar), without cookies;
• title, description, canonical, Open Graph tags and robots meta tags;
• links: the address, the link text (up to 100 characters) and the nofollow flag;
• forms (the method, the site they submit to, and the field types, without anything typed in), frames and the addresses of external scripts;
• the beginning of the scripts embedded in the page (up to 300 characters each) and the beginning of the visible page text (up to 500 characters);
• the robots.txt file, the sitemap and the certificate expiry date.

Phone numbers and email addresses in the recorded texts and addresses are masked. Links to a phone, an email and messenger numbers (tel:, mailto:, WhatsApp, Viber) are recorded only as a kind of link: the number or address itself is not stored. From this the service builds a report on what keeps the site from being found and from bringing in customers. The speed of the home page is measured separately by Google at the service's request; that measurement is made by Google, not by our robot.

How it behaves

• It first checks the site's certificate: it opens a secure connection to the main address and reads the expiry date. No pages are requested at this point.
• Before loading pages, it reads robots.txt and follows it.
• It requests only public addresses of your site, with the GET method. It does not run scripts, sign in to accounts, fill in forms or send cookies.
• It pauses at least 1 second between requests to one site. It treats an address with www and without it as different sites, each with its own pause.
• It takes no more than 25 HTML pages per check.
• The result of a check is kept for 24 hours, so repeated checks of the same site do not load it again. The robot itself keeps the crawl records for 7 days after a check. The report stays in the account of the person who ordered it.

How to block it

Add a group for our robot to robots.txt at the root of your site (https://your-site/robots.txt):

User-agent: PrivlekaiBot
Disallow: /

Then the robot will not load any page except robots.txt itself. You can also close only a part of the site: Disallow: /private/. The robot understands Allow, Disallow, and * and $ in rules, as described in RFC 9309. The name in the User-agent line is PrivlekaiBot; the case does not matter.

If there is no group for PrivlekaiBot, the robot follows the User-agent: * group. On any 4xx response (except 429) in place of robots.txt, it assumes there are no restrictions. If the file cannot be obtained (a network error, a 5xx or 429 response, a redirect that never ends), the robot loads nothing except the file itself. The robot ignores the Crawl-delay line: its request rate is limited anyway.

Contact us

Does the robot behave differently from what is described here, or do you want it to stop visiting while you cannot change robots.txt? Write to softispower@gmail.com with the site address and the visit times from your log, and we will look into it.