KlasprBot, and how to stop it.

1. What KlasprBot is

KlasprBot is the crawler behind klaspr, a competitive-intelligence tool. When one of our customers tracks a company, KlasprBot reads a small number of that company’s public pages and records what they said and when — so that a claim like “this plan went from €19 to €29” can be traced back to the page it came from.

It identifies itself honestly on every request, and the identifier exists so that a site which does not want to be read can block it. That is the point of sending it.

2. How it behaves

User agent

KlasprBot/1.0 (+https://klaspr.com/bot)

Requests per page

One GET per reading. No parallel connections to the same host.

Frequency

At most once per day per monitored page, in a single daily pass.

Timeout

15 seconds. A slower response is abandoned, not retried in a loop.

Size limit

5 MB. Larger responses are dropped without being read.

Accepted content

HTML only. Any other content type is discarded on arrival.

Requests carry an Accept-Language header matching the market being observed, and ask for an uncached response. Both exist to avoid comparing two different versions of your page and reporting the difference as a change you made.

3. What it reads, and what it never does

KlasprBot fetches business pages: a homepage, a pricing or plans page, a product page, a changelog. It finds them by trying conventional paths such as /pricing, /plans or /tarifs.

It does not, in any circumstance:

• create an account, sign in, or read anything behind authentication;
• submit a form, add to a cart, or trigger any state change;
• bypass a block, rotate user agents, or disguise itself as a browser;
• collect names, email addresses, CVs, employee profiles or any personal contact detail;
• execute your JavaScript or load your ads, images or trackers;
• crawl your whole site — it reads specific pages, not a domain.

4. robots.txt

KlasprBot does not currently read robots.txt. We would rather state that plainly than let you assume a directive is being honoured when no code reads it. Support for it is planned and is not built.

Until it is, a Disallow rule aimed at KlasprBot will have no effect, and the two methods in the next section are the ones that work. Both are honoured immediately and neither requires our cooperation.

5. How to block it

Block by user agent at your edge. Any CDN, WAF or server rule matching the string KlasprBot will stop it. KlasprBot treats any non-successful response as “no observation”: it records nothing, produces no comparison, and the source is marked unreachable. Returning 403 or 429 is enough — there is no fallback path and no retry storm.

Ask us. Write to [crawler contact email] from an address at the domain concerned, or with something else that establishes you speak for it, and we will stop reading it. Tell us the domain; you do not have to explain why.

Blocking KlasprBot has no effect on how your company is described by anyone else, and no effect on your visibility anywhere. There is no downside we are aware of, and no incentive we have to make it awkward.

6. What is stored, and for how long

For each reading, klaspr keeps the page content as retrieved, the URL, the exact time of the reading, and the structured data extracted from it. That copy is the evidence behind every statement the product makes about your company — a claim shown without the page that supports it is exactly what klaspr is built not to do.

Stored readings are append-only: a past observation is never rewritten or backdated. No automatic deletion is in place today — a ninety-day purge of stored page content is planned and is not built. Deletion on request is available now, see below.

7. Corrections and removal

If klaspr holds something about your company that is wrong, or that you want removed, write to [crawler contact email] with the domain and, if you can, the specific page or statement. We will correct or delete it and stop future readings if you ask us to.

If you are a sole trader or freelancer and the information identifies you personally, this is also a right you hold under the GDPR rather than a favour we grant. The privacy policy sets out those rights and the deadlines that apply to them.

8. Verifying it was really us

Anyone can put our string in a user agent header, and impersonating a named crawler is a common way to make abusive traffic look legitimate. KlasprBot does not yet publish a list of source addresses or support reverse-DNS verification, so a request claiming to be KlasprBot cannot currently be proven to be one.

If you are seeing traffic that claims to be KlasprBot and does not match the behaviour described in section 2 — many requests per second, full-site crawling, form submissions, attempts to sign in — it is almost certainly not us. Send us the logs at [crawler contact email] and block it.