Claude-User

What Claude-User does

Claude-User fetches individual pages on behalf of people using Claude, distinct from ClaudeBot's bulk training crawls. Anthropic documents separate robots tokens for training, search, and user-initiated fetching so site owners can treat each differently.

Why it visits your site

A person asked Claude something your page answers. Single-page, on-demand fetches are the pattern.

The bigger picture: AI user agents

User-agent fetchers are not crawlers in any meaningful sense: each request happens because a specific person asked an assistant about a specific page, right now. The pattern in your logs is the opposite of a crawl - single pages, irregular timing, no systematic walk of your archive. It is closer to a human visit with an unusual browser than to bot traffic, and treating its volume as an engagement signal - which of your pages do people ask AI about? - is more useful than treating it as load.

This category is also where the robots.txt social contract gets genuinely contested. Some operators apply robots rules to these fetches anyway; at least one documents that it does not, on the argument that an agent acting on a user's explicit request is the user's browser, not a robot. Wherever that debate settles, the practical takeaway for site owners is that policy files govern this category unevenly, and server-side visibility is how you know what is actually happening.

Should you block it?

Same logic as ChatGPT-User: blocking it hides you from users at their moment of interest while saving you almost nothing in load. Allow it unless you have a strict no-AI policy.

Block via robots.txt

Add these lines to the robots.txt at your site root to ask Claude-User to stay away:

User-agent: Claude-User
Disallow: /

Remember that robots.txt is a request, not a lock - compliance is voluntary, and enforcement happens at the server. Viz's AI Control pairs robots.txt rules with server-level 403 responses and then probes your site to verify the block is actually holding.

How to see Claude-User traffic on your site

You have three windows onto it, each with a catch. Raw hosting access logs contain every Claude-User request, but many managed WordPress hosts don't expose them, and when they do you are grepping text files by hand. A CDN dashboard sees the traffic at the edge, but it lives outside WordPress and speaks in totals, not in your site's terms. And JavaScript analytics - GA4 and friends - will never show it at all, because Claude-User doesn't run scripts.

Viz's answer is the request log: filter to Claude-User and read exactly which URLs it fetched and when, right inside wp-admin - then watch the same filter after any blocking decision to see whether the visits stopped, slowed, or kept coming.