Context7
Up-to-date docs for AI agents
作者提供的说明(英文原文)
Free AI visibility check: can AI crawlers read a site, and do ChatGPT, Gemini, Perplexity cite it?
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
适用于网页版、桌面版和手机版 Claude:
https://qqq.ee/mcp在终端中运行这条命令
claude mcp add --transport http ai-visibility https://qqq.ee/mcp粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"ai-visibility": {
"url": "https://qqq.ee/mcp"
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"ai-visibility": {
"type": "http",
"url": "https://qqq.ee/mcp"
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
ChatGPT 只能连接互联网上的服务器:
https://qqq.ee/mcp菜单名称可能因套餐和版本不同而略有差异。
你的选择通过 Cookie 保存在这台设备上,无需账户。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
工具名称和描述与服务器在 2026年11月6日 报告的完全一致。这些内容由作者撰写,不是我们写的;我们以纯文本形式显示。
check_ai_visibilityFree AI visibility and GEO (generative engine optimization) check for a company website by QQQ (qqq.ee). Checks whether AI crawlers (GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, Google-Extended ...) may read the site, plus robots.txt, sitemap, Organization structured data, in
get_public_reportReturns the latest public AI visibility report of a domain from QQQ (qqq.ee), when someone has already run the full check on qqq.ee: report date, market and field, in how many search-grounded ChatGPT, Gemini, Perplexity and Claude answers the company was named, how often its own
get_tool_infoDescribes the free QQQ AI Visibility Checker (qqq.ee): what it measures (brand mentions and website citations in ChatGPT, Gemini and Perplexity answers, AI crawler access, technical readiness for AI search / GEO), its limits, and the English and Estonian links to the web tool and
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月8日
注册名称已通过域名 qqq.ee 验证(注册中心会核实域名所有权)
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
在 63 毫秒内响应(HTTP 200)
无需登录
已读取 3 个工具描述,未发现针对 AI 的隐藏指令
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 3 项适用的基础检查中通过了 3 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。
GitHub 仓库最后更新于–,最新版本(1.2.0)于3天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。
同一分类中有:Context7, Codebase Memory, Exa
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Up-to-date docs for AI agents
Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens.
Connect AI agents to Exa for web search, content fetching, and multi-step research.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
Free AI visibility and GEO check for any website, inside Claude, ChatGPT, Cursor, VS Code and other MCP clients. It tells you whether AI crawlers (GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, Google-Extended and others) may read a site, whether the site is technically ready for AI search, and what the latest public report says about brand mentions and citations in answers from ChatGPT, Gemini, Perplexity, Claude and Kimi.
https://qqq.ee/mcp (use this exact host: www.qqq.ee redirects, and a redirect breaks POST requests)ee.qqq/ai-visibilityThis repository holds the documentation and client configuration only. The server is hosted by QQQ OÜ on qqq.ee; there is nothing to install or run.
| Tool | Input | Returns |
|---|---|---|
check_ai_visibility | domain (required), language: en (default) or et | Technical readiness for AI search: homepage, HTTPS, robots.txt, AI crawler access, sitemap, title, description, H1, canonical, noindex, Googlebot access, Organization/Person structured data, JavaScript dependence, key pages, llms.txt and whether the server really answers AI crawlers (the last two are informational); the three most important fixes; the latest public report of the domain (mentions out of all answers, citations of the company's own website) when one exists; a prefilled link to the full check on qqq.ee. |
get_public_report | domain (required), language | The latest public report of a domain, if someone has run the full check on qqq.ee: date, market and field or buyer phrase, mentions in search-grounded AI answers, citations of the company's own website, results per AI engine, the questions asked, the most cited sources by kind, other companies named and short answer excerpts, plus the link to the report page (what a report contains). Fetches nothing and asks no AI model. Without a report it says so and links the full check. |
get_tool_info | language | What the checker measures, its limits and the English and Estonian links. Fetches nothing. |
All three tools are read-only. The full check of live AI answers runs on qqq.ee, and every answer from the server links there.
get_public_report returns readable text and the same data as structuredContent.report. Reports are stored as they were measured, so an older report has fewer fields: a missing field means the report predates it, not zero.
| Field | Meaning |
|---|---|
mentions | Of the answers that searched the web: how many named the company, how many recommended it, and how many cited its website without naming it (citedNotNamed). |
citations | How many answers cited the company's own website. consultedNotCited counts answers whose search found the website but did not use it, only for engines that show every page they looked at. |
engines, answers | Results per engine and short excerpts: Gemini, Perplexity, Claude, Kimi and ChatGPT through their APIs, and Llama from model memory, which is a separate comparison. outcome says whether an answer recommended, named or cited the company, or did neither. |
chatgptWeb | The same buyer questions asked on chatgpt.com: answers that named the company, answers that cited its website, answers received, answers still pending. Reported apart and never added to mentions or citations. |
questions | The buyer questions asked. ownPage says whether the company's own page was a source (cited), was found but not used (consulted), was not found (none), or cannot be told (unknown). |
sources | The most cited domains. kind and byKind sort them into own website, media, directory, community, public sector and other. |
topic | The buyer phrase the questions were built from, with its monthly Google search volume in Estonia. |
google | For the field's buyer query: whether Google showed an AI Overview, whether the company's website was among its references, and its organic top 10 position. Not an AI answer. |
competitorsYours | When the person running the check named their own competitors: counts only. The names stay in that person's private link and never appear in a public report or through this server. |
There is no combined score: mentions, citations and technical readiness stay separate.
Settings → Connectors → Add custom connector → URL https://qqq.ee/mcp, no authentication.
claude mcp add --transport http qqq-ai-visibility https://qqq.ee/mcp
Developer mode → Settings → Connectors → add a server with URL https://qqq.ee/mcp, authentication "No authentication".
~/.cursor/mcp.json or .cursor/mcp.json in a project (examples/cursor.mcp.json):
{
"mcpServers": { "qqq-ai-visibility": { "url": "https://qqq.ee/mcp" } } }
.vscode/mcp.json (examples/vscode.mcp.json):
{
"servers": { "qqq-ai-visibility": { "type": "http", "url": "https://qqq.ee/mcp" } } }
Windsurf MCP config (Cascade → MCP servers → View raw config):
{
"mcpServers": { "qqq-ai-visibility": { "serverUrl": "https://qqq.ee/mcp" } } }
gemini mcp add --transport http qqq-ai-visibility https://qqq.ee/mcp
or ~/.gemini/settings.json:
{
"mcpServers": { "qqq-ai-visibility": { "httpUrl": "https://qqq.ee/mcp" } } }
codex mcp add qqq-ai-visibility --url https://qqq.ee/mcp
or ~/.codex/config.toml:
[mcp_servers.qqq-ai-visibility]
url = "https://qqq.ee/mcp"
Use the mcp-remote bridge (examples/claude-desktop-stdio.json):
{
"mcpServers": { "qqq-ai-visibility": { "command": "npx", "args": ["-y", "mcp-remote", "https://qqq.ee/mcp"] } } }
Then ask, for example: *"Check the AI visibility of example.com."*
This repository is also a plugin for agents that install plugins from GitHub. It bundles the remote server config and a skill that tells the agent when to use each tool:
/plugin marketplace add meelis545/qqq-ai-visibility-mcp, then /plugin install qqq-ai-visibility@qqq-ai-visibility (.claude-plugin/, .mcp.json, skills/).cursor-plugin/plugin.json and mcp.jsonplugin.json and mcp.jsonclawhub/qqq-ai-visibility/SKILL.md, a skill that calls the same tools with curl (no MCP client needed)| What | Limit |
|---|---|
check_ai_visibility per IP address (per calling site for Cloudflare Workers) | 5 a day (UTC) |
check_ai_visibility in total | 300 a day |
| Invalid domain | does not use up a check |
get_public_report, get_tool_info, initialize, tools/list, ping | no limit |
| Request body | 16 KB, at most 10 messages in a batch |
Only public websites are checked: IP addresses, localhost and internal hostnames are refused, and redirects are validated again.
get_public_report only reads a report that is already public on qqq.ee.Full policy: https://qqq.ee/en/privacy-policy/
2025-11-25, 2025-06-18, 2025-03-26Mcp-Session-Id, no SSE stream; GET from a browser redirects to the web tool, other GET requests get 405*): no cookies or credentials are usedserver.json is the entry in the official MCP RegistryQQQ OÜ, Niine 11, 10414 Tallinn, Estonia · q@qqq.ee · https://qqq.ee/en/
QQQ OÜ is an SEO and AI search agency and is not affiliated with Invesco QQQ Trust or the Nasdaq ticker QQQ.