# Reading Perplexity requests on a WordPress site

- Author: REBASE-SG
- Published: 2026-11-01 09:56:30
- Modified: 2026-11-01 09:56:30
- Canonical URL: https://rebase-sg.com/perplexity-request-roles-wordpress-access-en/

---

A request carrying the Perplexity name appears in a site log. What does it establish? It may relate to search discovery, or it may be a page fetch prompted by someone using the service. Combining those roles into one “AI crawler visit” removes context before the site owner starts interpreting indexing, citations or traffic.



This is practical analysis based on current documentation, not a report of a feature launched today. Perplexity’s crawler documentation distinguishes PerplexityBot, which supports surfacing and linking websites in search results, from Perplexity-User, which supports user-requested access. It also says that the latter generally ignores robots.txt because a user requested the fetch.



The original illustration above shows those roles side by side: search discovery and a user-requested fetch. It contains neither private logs nor results from a crawling test of this website.



Start with what the request received



For a WordPress owner, a useful investigation asks which kind of request reached which public URL, at what time, and what response it received. If someone wants an assistant to read an installation guide but the fetch receives an error or a verification page, improving the guide’s wording will not resolve that particular access failure.



Opening the page successfully in a browser is useful context, but it does not establish what another client received. Requests can encounter different cache states, rules or session conditions. Keep the investigation attached to the reported request rather than concluding that a successful browser visit proves the problem belongs to the other platform.



Choose a guide that is genuinely intended for public reading. Record its exact URL and the time of the reported problem, then look for corresponding server or security events. Keep the response status, any matching rule, and the result of source verification. If there is no corresponding event, preserve that gap instead of filling it with an assumed cause.



Identity deserves attention as well. The Perplexity documentation publishes separate address sources for the two roles and combines the appropriate User-Agent with its address range in WAF examples. A recognizable header alone should not be treated as verified identity. Consult the current official sources when configuring a system rather than copying an address list from a blog post that may become stale.



This makes an operational difference when discussing a problem with a hosting provider. A URL, timestamp and observed response give support a concrete event to investigate. Saying that “AI cannot read the website” leaves open which client, page and request path were involved. A narrowly described failure is easier to diagnose.



Crawl preferences and access authorization have different jobs



RFC 9309 states that robots.txt rules are not access authorization. They express crawling instructions; they do not replace the controls that determine whether a requester is entitled to read confidential material.



For a public tutorial, identify the failing layer before changing policy. If an event shows that a specific rule blocked a verified request the owner wants to permit, assess the scope of that rule. A problem affecting one URL is insufficient reason to change protections across the whole site. Successful access would still be a different outcome from being selected as a cited source.



For a page containing client information, decide who is authorized to read it and enforce that decision through authentication and authorization. Relying on a fetcher to honor a robots preference would leave the underlying access question unresolved. Public documentation and confidential reports need different decisions.



Keep the conclusion of an investigation as precise as the evidence allows: a verified request reached this URL at this time and received this response. That can support a maintenance decision. Whether the page later appears as a source, attracts a reader or contributes to an installation requires separate observation. None of those outcomes follows just from the name attached to the request.



中文版：WordPress 日志里的 Perplexity 请求，先分清它来做什么。