ItsMyIp

Test a site's robots.txt

Who is allowed to crawl your pages? The verdict bot by bot — search engines, AI crawlers, link-preview bots — with the exact rule that applies.

01What is this tester for?

The robots.txt file tells bots what they may crawl. One badly written rule and the whole site vanishes from Google — or, increasingly, from the answers of ChatGPT, Claude and Perplexity: many sites block AI crawlers without knowing it, through a generic rule or a copy-pasted firewall setting. This tool reads the file live and returns the verdict bot by bot, with the exact rule that applies.

The subtlety lies in the priorities: each bot picks the most specific User-agent group that concerns it, then the longest rule wins, with Allow prevailing on ties. A Disallow: / under User-agent: * therefore does NOT block Googlebot if a User-agent: Googlebot group exists above. Pair this with our page audit, which also checks robots.txt and sitemap.

02Understanding the directives

User-agent
Opens a rule group for one bot (User-agent: Googlebot) or all (User-agent: *). A bot obeys a single group: the most specific one for it.
Disallow
Forbids paths starting with this prefix. Disallow: / blocks everything; an empty value forbids nothing.
Allow
Re-allows a path inside a forbidden area (Allow: /blog under Disallow: /). At equal length, Allow beats Disallow.
* and $
* matches any sequence of characters, $ anchors the end (Disallow: /*.pdf$ blocks PDFs only).
Sitemap
Declares the XML sitemap's URL — the only directive read outside any group, by all bots.
Crawl-delay
Request spacing asked of some bots. Ignored by Google; Bing honours it.

03Block or allow AI crawlers?

Blocking GPTBot, ClaudeBot or Google-Extended protects your content from model training — but also removes your pages from AI assistants' answers, a growing share of site discovery. Note the difference between training bots (GPTBot, ClaudeBot) and search or user-action bots (OAI-SearchBot, ChatGPT-User, Claude-User): you can block the former and allow the latter.

Whatever your choice, make it consciously: check here what your file actually says, and keep in mind that robots.txt is a convention, not a barrier — a misbehaved bot can ignore it. For actual blocking, act at the server or firewall level (WAF rules, AS blocking).

04Frequently asked questions

No robots.txt: is that a problem?

No. Without a robots.txt (404 response), all bots consider everything allowed — a perfectly valid choice for a fully public site. The file becomes useful when you want to exclude areas (admin, internal search, carts), declare your sitemap, or control AI bots. However, a robots.txt answering with a 5xx error is a real problem: Google pauses crawling the site.

Does robots.txt keep a page out of Google?

Not exactly: it prevents crawling, not indexing. A page blocked by robots.txt can still appear in results (without a description) if other sites link to it. To exclude a page from the index you need a meta robots noindex tag — and thus, paradoxically, to allow the bot to visit the page so it can see it.

What is the difference between GPTBot, OAI-SearchBot and ChatGPT-User?

Three distinct uses at OpenAI: GPTBot collects for training future models; OAI-SearchBot indexes for ChatGPT's built-in search; ChatGPT-User visits a page at a user's request (when they paste your URL into the conversation). You can block training while staying visible in search: Disallow for GPTBot, Allow for the other two. Anthropic offers the same granularity with ClaudeBot and Claude-User.

My robots.txt blocks a page, but it is already in Google. What now?

The robots.txt block removes nothing from the index — it even prevents Google from coming back to see the noindex tag you may have added. The correct sequence: unblock the path in robots.txt, add noindex to the page, wait for de-indexing, then re-block if needed. In an emergency, Search Console's removal tool takes the URL down within hours (temporarily).

Are the rules case-sensitive?

Paths, yes: Disallow: /Admin does not block /admin. Bot names, no: user-agent: googlebot and User-agent: Googlebot are equivalent. Also mind URL encoding: the rules apply to the path as it appears in requests, parameters included if you mention them.

05More tools