The front page tells you what this is. This page shows you how it thinks — the method, the framework, the catalog, and what your research becomes once you have it.
Every scan — a hundred-plus of them, across twelve domains — runs the same three-step research discipline. The steps are why the output is trustworthy.
Bring whatever you have — a headline, an entity, a scenario, a pasted URL, or a plain question. Each scan applies a specific analytical method to it, rather than whatever a chat model felt like producing that day. Not sure which scan? The Router reads your question and picks one; the Planner designs a linked sequence toward a goal.
The scan runs against live search and domain-specific data — SEC EDGAR for finance, FRED for economics, GDELT for defense, NASA EONET and NOAA’s CO2 record for climate, ESPN scores, standings and injury reports for sports, OFAC for sanctions screening, and more. The grounding constrains what the research can claim: quantitative statements are anchored in time, names and titles are checked, and what can’t be verified is flagged rather than smoothed over.
What comes back is not a wall of text. It’s finished research with named fields — what happened, what it means, who’s exposed, what to watch — sized by the FLOW framework and ending in the sources it consulted, each one a click from the claim it supports. Structured output can be handed to a board, a class, or a teammate without translation.
Two axes decide everything: SCALE — the organizational weight of a situation (dollars, authority, stakeholders) — and COMPLEXITY — how hard the right call is to see and to execute. Where a situation lands tells you what to do with it: delegate, resource, analyze, or mobilize.
One scale, three voices. The same letter answers a different question depending on the scan. On an Event or Entity scan it sizes a connection you’re interpreting; on a Risk Scan, Action Plan, or Options it sizes an action you’d be executing; everywhere, it sizes how much of your attention the thing deserves.
A FLOW C cannot exist at large scale. High complexity at large scale is a D — the scale decides.
At large scale, scale alone drives the classification. You don’t debate the difficulty of a billion-dollar problem before escalating it.
A and B differ only in scale. Both are low complexity — the question is just how much weight rides on the routine call.
One vocabulary across every domain — a FLOW D in finance reads the same as a FLOW D in policy. That’s the point: the classification travels, so your judgment does too.
Six ways in, then the rest of the catalog organised by the move you’re making — establish the background, map the argument, work the record, analyse the connections, decide what to do. This is the Home domain in full; every other domain carries its own set, renamed in its own language — each has its own page, below.
Don’t know which one you need? The Router reads a plain question and picks one; the Planner takes a goal and designs a linked sequence. Every completed scan also opens onto its own follow-ups — ask it a question, re-check what has changed since, red-team its judgments, or push a signal onto your watchboard.
The same grading, the same stamps, the same exports, whatever the subject. Each domain renames the catalog in its own language and adds the objects only it has — and each has its own page: who comes there, what grounds the research, how FLOW reads it, and the catalog in full.
A scan isn’t trapped on the screen you ran it on. Every completed scan converts to the shape the moment needs.
These are the positioning questions. The practical ones — pricing, credits, sharing, what the grades mean — live on the full FAQ.
The preview is the real product with real research — no account, no card. Pricing, when you want it, is one page over.