How to Research a Crypto Contract Before You Touch It
// The house research order — three reads, cheapest first, before any strategy page's mechanics even apply. It produces information, not a decision.
Three reads, cheapest first — and none of them is a decision. This is the order experienced users check a contract in, and what each surface actually displays.
// WHAT THIS IS
The funnel is the house research order: three reads, cheapest first, before any strategy page's mechanics even apply. It produces information, not a decision. Nothing on this page tells you to buy, short, or enter anything — it shows what each surface displays and the order experienced users check them in.
What THE FRONT and THE TURN mean, and what a confirmed flip is.
// STEP 1 — FIND IT ON SPOT CARRY
The board is titled SPOT CARRY, and its own subtitle says what it is: "// Buy spot, short perp — delta-neutral funding yield." It is not a price-lookup tab. Its search reads "Search spot strategies by symbol..." — you are searching STRATEGIES, and what comes back is a priced two-leg structure, not a quote. That is why it is step one: before any strategy page's mechanics apply, this board tells you whether the asset carries a structure here at all, and on what terms. The card states both legs and their cost: SHORT PERP · LONG SPOT · GROSS · COST / BE, with NET APR on the header beside the symbol. Read COST / BE before the headline rate — it is the toll the structure pays before it earns anything.
// OPEN SPOT CARRY
// STEP 2 — READ THE TREND STATE ON WAR TRENDS
The board shows the current trend read — one of the board's declared states; SIDEWAYS is what most rows print in this page's example — alongside two prices that are not the same thing: THE FRONT, the live trailing line the read invalidates through, and THE TURN, the last confirmed flip stamped with how long ago it fired (and absent when no flip is in window). What a confirmed flip is: the anchor page owns that definition — link, don't re-teach. The read is a description of where price sits relative to THE FRONT. It is not a forecast.
// READ THE TREND
// STEP 3 — RUN STEEL CHECK ON THE CONTRACT
STEEL CHECK inspects the contract itself and displays a verdict. There are four, and these are the words the panel prints: CLEAN STEEL · RUST DETECTED · TRAPPED — DO NOT TOUCH · // NO DATA — UNVERIFIED. Six named reads sit beneath the verdict, each carrying its own pill: CLEAR · RUST · DANGER · // NO DATA. The panel states its own limits, and this page repeats them rather than softening them: "STEEL CHECK reads the contract's on-chain properties — a safety scanner is not infallible; not a guarantee or advice," and "A point-in-time read — mutable contracts can change after you check." A clean verdict is a screening result, not a safety guarantee; a flagged verdict is a displayed finding, not an accusation. Both sentences appear next to the plate at equal weight.
// RUN STEEL CHECK
// THE WORKED PASS (WHAT THE PLATES ON THIS PAGE SHOW)
Read the three example plates in order and the funnel stops being abstract. The Step-1 plate is the SPOT CARRY board: each row pairs the two legs the board's own labels name — SHORT PERP and LONG SPOT — then prints a GROSS figure, the COST / BE column, and NET. The distance between GROSS and NET is the whole lesson of that board: the headline is not the take-home, and the board says so itself, row by row. The Step-2 plate is the trends board on a day when most rows print SIDEWAYS — which is the honest normal, not a broken screen. The Step-3 plate is a verdict panel reading CLEAN STEEL with its six pills reading CLEAR, and two disclaimers the product locks onto the panel. Three screens, three different kinds of fact: what a position pays, where price sits relative to a line, and what an inspection could see. None of them is an instruction.
// READING A CARRY ROW WITHOUT BEING SOLD BY IT
On the Step-1 board most eyes go straight to the biggest number, and the board's own layout is the antidote. SHORT PERP and LONG SPOT name the two legs — the structure described is delta-neutral, earning from the relationship between the legs rather than from price going up. GROSS is the pre-cost figure. COST / BE puts a number on what the position spends and where it breaks even. NET is what remains, and it is the only column that survives contact with reality. A row where GROSS is loud and NET is quiet is not a trap — it is the board doing its job in public. Any rate on any row is a snapshot of that moment's funding, which changes constantly; the capture band on the plate says so on the image itself.
// WHEN THE FUNNEL SAYS NOTHING
A funnel pass has more than one honest ending. The asset may not appear on the boards at all. The trend row may print SIDEWAYS — on the day this page's examples were captured, most rows did — which is the board declining to claim a direction, not hiding one. The verdict panel may come back // NO DATA — UNVERIFIED, which is an absence of evidence and gets treated as exactly that. In all three cases the pass produced its output: nothing here compels an action, and "no read" is a complete, valid result. The most common mistake with a research tool is treating a neutral screen as a puzzle to push through rather than an answer.
// A VERDICT IS A SNAPSHOT
Every figure the funnel touches is time-stamped by nature. Carry figures move with funding and price; a trend read holds only until the next candle argues; a verdict panel reports what the check could see at the moment it ran. That is why the screenshots on this page carry their capture band, and why a pass through the funnel ages the way any snapshot does. Re-running the three reads costs minutes; assuming yesterday's pass still holds costs whatever the market decides it costs. The disclaimers the product locks onto its panels exist for the same reason.
// WHY THE ORDER IS CHEAPEST-FIRST
The sequence is not decoration. Step 1 is a table read — seconds, no judgment required. Step 2 adds one interpretive layer: a state and a line, both defined once on the anchor page. Step 3 is the only step that examines a specific contract. Running them in this order means a pass can end early at the cheapest possible point: an asset that never appears on the boards, or a row that prints no direction, ends the pass before the contract inspection was ever needed. Reversing the order spends the most careful read first on candidates the cheaper reads would have filtered. The funnel's value is not any single screen — it is that the screens are ordered.
// WHERE THE FUNNEL HANDS OFF
The funnel ends where the strategy pages begin. A pass that survives all three reads has produced exactly one thing: an asset that exists on the boards, a stated trend context, and an inspected contract. What to do with that — if anything — is the territory of the strategy pages, each of which opens with its own risk section for a reason. The funnel filters unforced errors out; it does not filter good trades in. Nothing about a clean pass makes the next step safe. It makes the next step informed.
// WHAT THIS FUNNEL DOES NOT DO
- It does not rank, score, or recommend assets.
- It does not predict. Every read above is current-state telemetry.
- It does not replace the strategy pages' risk sections — it runs BEFORE them.
- A full pass through all three steps can still precede a losing position. The funnel reduces unforced errors; it does not create edge.
