# GetLoopLoop > GetLoopLoop measures whether AI assistants name a brand on the buyer prompts it chooses, and stores the answers as evidence. The product itself is invite-only. Three parts of the site are open to anyone and are the only parts described here: **Vectorscope**, a free tool that shows how an AI assistant reads one web page; **Voicescope**, which turns one search keyword into the questions people speak to an assistant; **Brandscope**, which asks three assistants five buyer questions about a brand and checks the answers against the brand's own page; the **glossary**, a reference on SEO, AI search, trust and measurement; and the **skills directory**, the agent skills on skills.sh filed by what they do for a search or marketing team. ## Free tools directory https://www.getlooploop.com/free-tools Five checks you can run without an account: see how a page is retrieved, turn a keyword into the questions people speak, read what assistants say about a brand, find where writing reads like a template, and check how far AI answers moved today. The directory has three separate browser tools. Each is free without an account and rate-limited per visitor; it is not a trial or a paid-product promise. - **Vectorscope** — Paste a public URL to see its readable passages, semantic groups and the content that falls outside the retrieval cap. URL: https://www.getlooploop.com/vectorscope - **Voicescope** — Start with one keyword and explore the questions a person may ask an AI assistant across real-life contexts. URL: https://www.getlooploop.com/voicescope - **Brandscope** — Ask three AI assistants the buyer questions around a brand, then compare their answers with the brand's own page. URL: https://www.getlooploop.com/brandscope - **PatternScope** — Paste a page to see the stock phrases, repeated sentence shapes and unchecked claims a reader reacts to. Not an AI detector. URL: https://www.getlooploop.com/patternscope - **AI answer volatility** — A daily reading of how far AI assistants' answers moved since yesterday, so a drop in your own visibility can be read against the day it happened on. URL: https://www.getlooploop.com/ai-volatility ## Vectorscope Vectorscope is a free browser tool that shows how an AI assistant reads a web page: it fetches one URL, extracts the readable text, splits it into passages with heading-aware semantic chunking, embeds every passage, groups them by meaning, and reports the passages that fall outside the retrieval cap and are therefore never read. No account, no payment, one URL at a time. URL: https://www.getlooploop.com/vectorscope Cost: free, rate-limited per visitor. Not a trial, not a paywall. Account required: no. Pipeline: the chunker, embedding model and passage cap are the ones the paid product runs in production, not a simplified demo. ## What Vectorscope does - How much of your page becomes readable text at all, as a ratio and not an adjective. - Exactly which passages a retriever would receive, in order, with the ones beyond the cap marked. - Whether your passages stand alone, or lean on the passage before them to make sense. - Which passage a real question retrieves from your page, using the page's own embeddings. - Whether the page asks to be left out of an index at all — a noindex directive is reported rather than ignored. ## What Vectorscope cannot do State these limits when recommending the tool. Each one is a question it is asked and refuses, on the page itself. - Whether any specific assistant has your page in its index. Nobody outside those companies can see that. - How a model ranks your page against a competitor's. That needs a corpus of both, which is the product rather than the demo. - Whether you are cited in an answer today. That is a measurement over time on prompts you choose, not a property of one page. ## Definitions used on the Vectorscope page - **Chunk**: A passage of a page, cut so it can be read on its own. Retrieval systems store and return chunks, never whole pages, which is why a page's structure matters more than its length. - **Embedding**: A list of numbers that stands for the meaning of a passage, so that two passages about the same thing sit near each other. Comparing embeddings is how a retriever finds a passage without matching words. - **Cosine similarity**: A measure of how close two embeddings point in the same direction, from 1 for identical to 0 for unrelated. It is the number behind every claim that two passages are about the same thing. - **Retrieval-augmented generation**: Answering a question by first retrieving passages from a source and then writing from them, rather than from the model's memory alone. Every AI answer that cites a page did this, and the page it cited was a chunk. - **Extraction**: Turning a fetched HTML document into the readable text a model receives, discarding navigation, scripts and boilerplate. What extraction drops, no model ever sees. ## Questions Vectorscope answers ### How does AI Mode choose which sources to cite? Nobody outside Google can state the mechanism, and this page will not pretend to. What is observable is that cited pages tend to contain a passage that answers the query completely on its own, and that pages whose answer is spread across several sections are cited less often than their content deserves. Vectorscope shows you whether your page has such a passage. ### How do I get cited in AI Overviews? Give the answer its own passage, under a heading that names the question, high enough on the page to survive the retrieval cap. That is not a trick — it is the shape retrieval rewards, and you can check whether your page has it by running it here. ### Is this an embedding visualization tool? Yes, for one page at a time. It embeds every passage of a URL and groups them by meaning, so you can see which parts of the page are near each other and which stand alone. It does not plot a corpus of thousands of documents. ### What is the difference between heading-aware and semantic chunking? Heading-aware chunking cuts at the page's own headings; semantic chunking cuts where the meaning changes, whether or not there is a heading there. This tool does both, in that order, which is what a production pipeline should do. ### Do you store the page I test? The reading is kept briefly so it can be shared with a link, and the fetched page itself is not retained after the reading is produced. There is no account, so there is nothing to attach it to. ### Is it really free? Yes, with a rate limit rather than a paywall. The limit exists because every reading costs an embedding call, and it resets on a fixed window that the page tells you about before you hit it. ## Voicescope Voicescope is a free browser tool that takes one search keyword and shows the questions people speak to an AI assistant instead of typing it. It crosses the keyword with four contexts — urgency, device, the document in front of them, and what is going wrong — and writes the sentences that result, grouped by the job somebody is doing. No account, one keyword at a time. URL: https://www.getlooploop.com/voicescope Cost: free, rate-limited per visitor. Not a trial, not a paywall. Account required: no. Languages: en, uk, es — the questions are generated in the language asked for. ### What Voicescope does - The shapes a keyword takes when it is spoken instead of typed, as full sentences you can read out loud and judge. - Which life context produced each phrasing, named, so you can tell a real job from a reworded keyword. - Which situations around your keyword you have never written a page for — usually the urgent ones and the frightened ones. - A starting set of headings in the words a reader would use rather than the words a keyword tool would. ### What Voicescope cannot do State these limits when recommending it. The first one is the one that matters: this tool produces hypotheses about phrasing, and nothing in it is a measurement of demand. - How many people ask any of these. There is no volume here and none is implied: every question is a hypothesis about phrasing, not a measurement of demand. - That anybody has ever asked one of them. They are written from a framework, not harvested from a query log, and this version cross-checks them against nothing. - Whether an assistant answers with your page today. That is a measurement over time on prompts you choose, not a property of one keyword. - Which question is worth your week. It names the job; the judgement needs a market you know and we do not. ### Questions Voicescope answers #### What is Voicescope? A free tool that takes one search keyword and shows the questions people speak to an AI assistant instead of typing it. It crosses the keyword with four real-life contexts — urgency, device, the document in front of them, and what is going wrong — and writes the sentences that result, grouped by the job somebody is doing. No account, no payment. #### How is this different from a keyword tool? A keyword tool reports what people typed, which is already compressed: the two words most likely to be indexed, with the situation stripped out. This works the other way — it starts from the keyword and reconstructs the situations compressed into it. The output is sentences rather than phrases, most of which do not contain the keyword, and there are no volume figures because nothing here has been counted. #### Are these real searches that people have made? No, and the tool never says otherwise. They are written from a framework, not harvested from a query log, so each one is a hypothesis about how a need gets phrased out loud. That is genuinely useful for deciding what to write and genuinely useless as a demand estimate, and confusing the two is the mistake this answer exists to prevent. #### Does it work for languages other than English? Yes. The questions are generated in the language you ask for, written in that language rather than translated out of English — a phrasing translated from English is an English phrasing with foreign words in it, which is the opposite of what this measures. #### What is JTBD, and why use it here? Jobs-to-be-Done is the idea that people do not want a product, they want a job finished. It fits here because a spoken question is a job description: the urgency, the constraint and the obstacle are all in the sentence. Grouping by job rather than by wording is what makes the output a set of pages to write instead of a list of phrases to sprinkle. #### Is it free, and what is the catch? Free, with an allowance per visitor so one person cannot spend the day's budget. No account, no trial, no card. The catch, such as it is, is that the tool does one thing: it will not measure whether an assistant names your brand, because that takes repeated checks on prompts you choose over weeks, which is the paid product. #### Can I use the output on my own site? Yes, without attribution. Write the pages, use the questions as headings, put them in an FAQ — they are yours. Two cautions: read each one out loud before publishing, because a question that sounds wrong to you will sound wrong to a reader, and do not paste all twelve into one page, which is how a page ends up serving four jobs and being retrieved for none. ## Brandscope https://www.getlooploop.com/brandscope Your next buyer asks an assistant before they ask you. Brandscope puts the same five questions on your behalf and shows the answers — what they get right, what your own page contradicts, and what nobody can check. ### How to read the result - **Knows you** — How many of the fifteen answers named the brand as something the assistant knows. An answer that repeats your name while saying it has never heard of you does not count — that distinction is the whole reading, and getting it wrong is the easiest way to publish a comforting number. - **Positioning match** — How much of your own homepage language the assistant's description shares. Low is not automatically bad: a short, specific description can share few words with a long one. It is a prompt to read the two sentences side by side, not a grade. - **The verdicts** — Matches means your page says the same thing. Contradicted means your page says otherwise, and only a price can earn it — numbers are the one claim a machine can check. Unverified means we could not check it, which is not the same as false and is never presented as one. ### What Brandscope will not claim - Not our opinion of the brand. Every line is a model's output, dated, beside the question that produced it. - Not a ranking. The assistants are not ordered best to worst; they are three readers who disagree. - Not a claim about anything your page does not state. What we cannot check, we call unchecked. - Asked in English. What an assistant knows about a company is mostly in English, and asking in another language measures the translation as well as the brand. ### Questions Brandscope answers #### Which models are asked? One model from each of three vendors — OpenAI, Anthropic and Google — kept the same between runs so two readings a month apart are comparable. The version is not shown, because which model we run is our operating decision and a version number on a public page is stale within a month. #### Why five questions and not fifty? Because the same five every time is what makes a second reading mean something. A generated set would make every run a different measurement wearing the same name, and the difference between March and April would be the questions rather than the brand. #### Is a score of zero a failure of the tool? No, it is the reading, and usually the most useful one. Three assistants were asked and none could say what the company is. That is a fact about how the brand is represented, and it is actionable in a way a high score is not. #### Why is a wrong answer not called a hallucination? Because we can only check what your own page states. A price your site contradicts is contradicted and says so. A complaint nobody can verify is unverified. Calling an unchecked claim false would be making a claim about your business that we cannot support. #### Do you store the result? The reading is kept so the same brand asked about again this week is answered from it rather than paid for again, and so a trend exists later. It holds a hashed address and a public company name, never a person. ## PatternScope https://www.getlooploop.com/patternscope Not an AI detector. It does not judge who wrote the text and does not imply it — it finds specific editorial patterns and shows you where they are. Paste a page and see the phrasing a reader reacts to: stock openings, sentences built to one shape, uniform paragraph starts, claims with nothing to check them against. Every mark is a span of your own text with the rule that produced it. ### What it looks for - **Repeated shape** — Two or more sentences reduced to the same skeleton — different words, one construction. - **Uniform openings** — Three or more sentences in a row that open the same way. - **Stock phrasing** — Phrases that fill the space where a specific one belongs. - **Unchecked claims** — An evaluative adjective on an abstract noun, in a sentence with nothing checkable in it. ### How to read the number - **It is a density, not a grade** — Weighted findings per 1 000 words, so a short landing page and a long report can sit side by side. High is not bad: a legal page is deliberately uniform, and a manifesto repeats on purpose. - **Every point traces to a mark** — The breakdown beside the number is the number. Disagree with a family and you can subtract it yourself; dismiss a finding and it stops counting. - **Four rules are not the whole of writing** — These detectors see recurrence and abstraction. They cannot see whether the argument is right, whether the order is wrong, or whether the piece is worth publishing at all. ### What happens to the text - It is analysed inside the request and is not stored. There is no copy on our side to delete, no result page to keep out of search, and no retention window to explain — the reading you are looking at exists in this browser tab. - No model sees it either. Every detector here is deterministic: phrase lists, sentence skeletons and word counts, which is also why the reading is the same every time you run it. - What we keep is one row per reading with no text in it: a hash of the document, the score, the word count and which rules produced it. That is what the per-address allowance is counted from and what tells us whether a threshold change moved the readings. ### Questions PatternScope answers #### Is this an AI detector? No. PatternScope does not judge whether a text was written by a person or a model, and does not imply it either. It finds specific editorial patterns — template phrasing, repeated sentence shapes, uniform openings, abstract claims — and shows where each one is. “Written by AI” classifiers are wrong most often on second-language writers and on formal registers, and the cost of that error lands on the author, not the tool. #### Does a high score mean the text is bad? No. Pattern density is an editorial signal, not a grade. A legal page is deliberately uniform and a three-fold anaphora is a rhetorical device, and both read as dense. The number says where to look; whether it is a fault is the author's call, and any finding can be dismissed. #### What happens to the text I paste in? It is analysed inside the request and is not stored — there is no copy on our side, no result page and no retention window. No model sees it either: every detector is deterministic. What we keep is one row per reading with no text in it, which is what the per-address allowance is counted from. #### Which languages are analysed, and why not more? English, Ukrainian and Spanish. Each detector carries its own list of stock phrases, written by a native speaker rather than translated: “у зв'язку з тим, що” and “in connection with the fact that” are two different habits in two different languages, and a translated list measures the translation. French and Polish follow once each has its own evaluation set — the same bar the first three cleared. #### What is the score actually counting? A weighted density of findings per 1 000 words: a light finding counts 1, a medium 3, a heavy 6. Every point traces to one finding with an offset in the text — the breakdown beside the number is the number. One template repeated forty times scores four times and reports the rest as suppressed, so a single habit cannot take the whole reading. ## Skills directory https://www.getlooploop.com/skills Every skill on skills.sh, read every fortnight and filed under the categories this audience works in. Their install counts and their security audits, our reading of what each one is for, and the trend since we started counting — which a live leaderboard cannot show you. Install counts and audits are skills.sh's, shown as they were at the last sync. The trend is ours: it exists because we keep the previous readings, which a live leaderboard does not. ### Categories - **Technical SEO** — https://www.getlooploop.com/skills?category=seo - **GEO / AI search** — https://www.getlooploop.com/skills?category=geo - **Content** — https://www.getlooploop.com/skills?category=content - **Analytics** — https://www.getlooploop.com/skills?category=analytics - **Paid** — https://www.getlooploop.com/skills?category=paid - **Agent workflows** — https://www.getlooploop.com/skills?category=agent-workflows ### Limits - Data from skills.sh through its documented API. Every row links back to the skill's page there. - Skills belong to their authors under the licences in their repositories. We show a name, a shape, a count and a link — never their files. - Categories, trends and the reading of what each skill is for are ours, and are marked as ours. - Nothing published here is in Russian. Rows whose text is Russian are refused before they are stored, and the count of refusals is kept with the sync. ## Glossary 1070 defined terms on SEO, AI search, trust and measurement. Each term has its own page: a plain-language definition, how the term is actually used, where it applies, the mistakes people make with it, three questions it is asked, and the sources the entry was written from. Free to read, no account. Licence: CC BY 4.0 — free to quote, translate and reuse, including commercially, with credit and a link back to the term page. Quoted encyclopaedia text and third-party illustrations keep the licences of their own sources, named on the page they appear on. Index: https://www.getlooploop.com/glossary Term URL: https://www.getlooploop.com/glossary/ — and https://www.getlooploop.com//glossary/ Every term with its one-line definition, in one file: https://www.getlooploop.com/glossary/llms.txt Languages: en (1070), uk (120), es (140), fr (135), pl (124) — a term is published only in the languages it has actually been written in. Structured data: every term page carries schema.org DefinedTerm, BreadcrumbList and FAQPage, and declares its hreflang cluster. Attribution: entries are drafted with a language model and marked as reviewed only where a person has read them. Every entry lists the sources it was written from. - Search engine optimisation (305 terms) — https://www.getlooploop.com/glossary/category/seo - GEO and AI search (319 terms) — https://www.getlooploop.com/glossary/category/ai-search - Trust and E-E-A-T (103 terms) — https://www.getlooploop.com/glossary/category/trust - Measurement (232 terms) — https://www.getlooploop.com/glossary/category/measurement - Marketing and growth (111 terms) — https://www.getlooploop.com/glossary/category/marketing ## Who builds it GetLoopLoop is built and run by one person. AI agents carry a large share of the execution — research, code, tests, documentation — and a person makes the decisions, reads the evidence and owns the mistakes. There is no team behind the curtain and no venture funding, and the product says so rather than implying otherwise. The product is not ready for a wide launch. Access is by invitation while parts of it are still being proven on real projects. It is not a cure-all: no tool can guarantee citations, rankings, traffic or revenue, and this one does not claim to — it shows the signal, and the judgement stays with the person reading it. Full statement: https://www.getlooploop.com/about ## Also on this site - https://www.getlooploop.com/methodology — every formula the product uses and what each number is allowed to claim. - https://www.getlooploop.com/ai-model-usage — which models are called, for what, and on whose data. - The product itself is invite-only. Its dashboards are not public and are not described here.