ArketicBot

01What ArketicBot is

ArketicBot is Arketic's web reader. It runs when an Arketic user asks it to read a page, when an organisation adds a site it chose to its knowledge base, or when a workflow crawls a site. It does not collect content to train AI models.

02Identities your server will see

Your access logs will show one of these User-Agent values:

• ArketicBot/1.0 (+https://arketic.ai/bot; info@arketic.ai)

  Page reading. A link shared in a chat, deep research, the site-reading block in a workflow, adding a company site to a knowledge base and news reading all use this identity.

• ArketicBot/1.0 (+https://arketic.ai/bot; info@arketic.ai) web-search

  Calls to the Wikipedia and Wikidata APIs during a web search.

• ArketicBot/1.0 (+https://arketic.ai/bot; info@arketic.ai) help-crawler

  Help-content crawler. It only visits Arketic's own domains (docs.arketic.ai, arketic.ai); it goes to no other site.

• Arketic-Workflow/1.0 (+https://arketic.ai/bot; info@arketic.ai)

  HTTP and integration blocks inside a workflow the customer built. This is not crawling; they are individual requests, like one from a browser or an API client.

The version number and the contact details in brackets can change between releases; the stable part is the name — ArketicBot and Arketic-Workflow. Match on the name when you search your logs.

03It follows your rules

robots.txt

it obeys the rules written for ArketicBot, otherwise those for * (RFC 9309). If your robots.txt answers with a server error (5xx) or cannot be reached, it reads nothing from your site; a 4xx means there is no robots.txt.

Parity with AI agents

if you block AI agents that read on a person's behalf (ChatGPT-User, Claude-User, Perplexity-User, OAI-SearchBot, Claude-SearchBot, PerplexityBot), ArketicBot does not read those paths either. Rules written only for training crawlers (GPTBot, ClaudeBot, CCBot, Google-Extended) do not bind ArketicBot, because Arketic trains no models.

Content-Signal

if you set Content-Signal: ai-input=no, your content is not given to AI as input and crawling of your site stops.

Politeness

it honours your Crawl-delay and makes one request at a time to a site.

One exception: workflow requests

Arketic-Workflow does not consult robots.txt, because it does not crawl: these are individual requests sent by a workflow the customer built — like a request from a browser or an API client. If you do not want them, block them by User-Agent on your server.

04To block it

add this to your robots.txt:

User-agent: ArketicBot
Disallow: /

This line stops every read that comes from ArketicBot. Workflow requests do not consult robots.txt (section 03), so they are not affected; block those by the Arketic-Workflow identity on your server.

05Contact

For questions or problems: info@arketic.ai