Media
3 items
Description
Check the current page's AI crawl access with robots.txt rules, live probes, and exportable reports with clear next steps. Can AI search crawlers access this page? GEO Check by Bee LLM explains the access signals it checks and helps you decide what to fix next. Open the extension on a website to check the current URL, including its path and query. Compare the robots.txt rules that apply to that URL with live requests made from your browser using AI crawler user-agents. See the result for ChatGPT, Claude, Perplexity and other crawlers, grouped by engine. Navigating to another page or changing the query uses that URL's check, rather than a result for the homepage. WHAT THE RESULT MEANS The toolbar icon stays neutral unless a completed, valid check is available for the current URL. While a check is running, an animated indicator shows that it is still working. The result appears after the check finishes; robots.txt rules alone do not produce an access verdict. A reused result is labeled as a saved check and shows when it was checked. Green, amber and red describe the access signals for the key AI search crawlers; gray means there is no confirmed result. Incomplete checks and uncertain responses are explained. A redirect does not confirm how the original URL responds to a crawler identity. A live probe comes from your browser and IP address. It does not come from OpenAI, Anthropic or another crawler operator. A website can treat verified crawler IPs differently. This is a technical access check, not a guarantee that an AI service will index, rank, recommend or cite your site. UNDERSTAND THE BLOCK Compare robots.txt restrictions with HTTP errors, challenge pages and rate limits. Reference and control requests help distinguish a crawler-specific response from a site outage or a connection-wide challenge. Where supported by the response, GEO Check identifies the CDN or protection service involved. Search, live assistant visits and model training are shown separately. Their crawler controls serve different purposes. Robots-only tokens and IP-verified crawlers are labeled accordingly instead of presenting an imitation request as a confirmed crawler visit. USE THE RESULT Read the summary first, expand crawler details, or rerun a check. Copy the complete report or use Export to download Markdown or JSON. Reports include the checked URL, check time and extension version, crawler rules and HTTP results, uncertainty, and suggested actions with verification steps. Query values, URL credentials and fragments are hidden in the displayed URL and exported reports by default; paths and parameter names remain visible. Reports do not include response bodies, complete response headers, session tokens or ChatGPT captures. Light and dark themes keep results readable. Five supporting page checks look at structured data, headings, viewport, dates and robots metadata. Their observations and suggested next steps appear separately from access results. Page signals are included in a report only when they match the checked URL. These are technical hints, not a citation score or a promise of visibility. OPTIONAL CHATGPT CAPTURE On chatgpt.com, choose to capture the conversation you have open. See the search queries and cited domains that ChatGPT makes available, and check whether your domain appears. No prompts are submitted and no conversations are generated. If full search metadata is unavailable, the extension can show the sources visible on screen and explains that limitation. This feature depends on ChatGPT's interface and may change when ChatGPT changes. LOCAL PROCESSING, EXPLAINED No Bee LLM account or subscription is required. Bee LLM does not receive your scan results or captured conversations. Results and preferences are stored in chrome.storage.local. The access cache uses the exact checked URL, including path and query values, to keep different pages and requests separate. URL redaction applies to display and reports, not to this local cache or the request sent to the tested website. Captured summaries can include titles, short prompt excerpts, search queries, citations and related metadata, and remain locally until cleared or replaced by newer captures. The ChatGPT session token is used in memory only to read the requested conversation from ChatGPT itself. The extension contacts the tested website. A cached user-agent diagnostic can also contact ifconfig.me, httpbingo.org, httpbin.org and postman-echo.com without sending the tested site's address or conversation data. These services see normal connection information such as your IP address and request headers. There is no analytics or advertising SDK in the extension. Settings lets you clear saved data. Built by Bee LLM: https://beellm.ai