> For the complete documentation index, see [llms.txt](https://docs.scanbylens.xyz/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.scanbylens.xyz/the-product/what-lensai-is.md). # What LensAI is The crypto AI that takes a stance and keeps score. LensAI is an automated crypto research desk. It reads the market the way a good analyst does, with many inputs considered together, and answers at three levels: * **The market as a whole.** Where each factor sits against its own history, which relationships have broken, and whether a move is spot-led or leverage-led. * **A single token in that context.** The token's own state, read against the market around it. * **Incidents.** When something breaks, a mechanism-level account of what happened: the trigger, the amplifier and the mechanism, held apart. Every read ends in **one call**, and every call is published to a permanent record that is scored over time. ## The principle: the join is the product Any capable language model can write fluently about markets. What is hard, and what compounds over time, is the evidence underneath: many inputs captured continuously, joined on one clock, normalized against their own history, connected by documented mechanisms, and checked before anything is said. LensAI's engine is that evidence layer. The model only narrates what the engine has measured. This is why LensAI can be blunt without being reckless: the bluntness comes from the evidence chain, not from a persona. ## Four assets that cannot be rebuilt later The outputs of LensAI are public and perishable by design: a call is published, and it is useful for days. The lasting value is in four assets that compound every day and cannot be assembled after the fact: | Asset | Why it cannot be rebuilt | | --------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | **The point-in-time corpus** | Every article and market observation is captured with the moment LensAI learned it. History seen from a later date is always contaminated by revisions; a record captured as it happened is not. | | **The normalized cross-factor history** | Years of joined, percentile-normalized factor states on one clock. Liquidations in particular are pushed live by exchanges and cannot be fetched again later. | | **The mechanism graph** | A curated, versioned map of how factors transmit to each other, with its full revision history. | | **The calibration record** | Every brief and every call, as made, with its outcome. A track record can only be built forward. | ## Product principles 1. **Show the work.** The evidence is the main object on screen, not a footnote. 2. **Blunt or silent.** One committed call, or an honest "no call". Never a pros-and-cons hedge. 3. **Numbers are real or absent.** Every figure is a stored row at the answer's anchor. Nothing is estimated to fill a gap. 4. **Spend model calls last.** Arithmetic and shared briefs first; language generation only where language is genuinely the job. 5. **The record is permanent.** Calls are never edited or deleted. Misses are shown first. ## What LensAI is not * **Not personal financial advice.** The same call goes to everyone. LensAI never tells a person what to do with their own holdings, never sizes a position, never sets an entry or a stop. * **Not a trading bot.** Analysis never connects to execution. There is no buy button, no copy-trading, no pre-filled orders. * **Not a signal service.** Calls are free and public, and they are scored for calibration, not advertised by win rate. * **Not a narrative machine.** Cycle theories and stories are context the system can reference, never patterns it claims to have found. * **Not a data race.** LensAI does not try to out-list the largest data vendors. It owns the column nobody fills: the join, the call and the record. ## Who it is for **Active traders**, retail and semi-professional, who check the market several times a day and want a fast, clear structural read, a committed call they can check later, and the evidence behind it without opening five dashboards. **Researchers, analysts and journalists** who come for the evidence, the sources, the mechanisms and the public record, to cite or to verify. **Builders and agents** that need verified, point-in-time market state through a tool interface rather than scraped text. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.scanbylens.xyz/the-product/what-lensai-is.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.