The whole book, ranked 24/7.
Hunter is GEM Hunter’s second perpetuals engine. It scans every linear perp on the venue, ranks heat, scores the top of the book against structure and liquidity, and prints a paper card only when the trigger chain confirms. Not a Telegram dump. Not a restyle of the five-profile desk.
Equal weight. Different physics.
The DEX scanner finds gems. The style desk trades liquid perps with independent books. Hunter hunts the whole tape so you are not stuck staring at BTC while the real move is three names down.
Five independent books
Scalp, Intraday, Swing, Position, Catalyst. Each has its own confirms, invalidation, and paper ledger. Styles never stack onto one candle. Regime on 15m / 1h / 4h / 1d is a filter, not a vibe.
- · FLAT microstructure (Funding, Liquidity/OI, Aggression, Time)
- · Macro blackouts per profile
- · Setup queue → live fill → history in R
Whole-market radar
A second engine. It ranks every perpetual on the venue around the clock, scores the top of the book against structure and liquidity, and prints only when the trigger chain confirms.
- · Own Telegram channel, own paper book, own bot toggle
- · Quality grades A+ / A / B — C never prints
- · Watch → Armed → Card. Heat is not a ticket.
Universe → heat → score → watch → armed → card.
Every stage can drop a name. That is how a 24/7 scanner stays quiet enough to read.
Every linear perp on the venue
Rank the whole book 24/7
Structure + liquidity composite
Clears the score bar
Sweep / HTF break in place
Trigger chain confirms — then print
What members see on /futures/hunter
- Live regime for the Hunter book, separate from the style-desk climate strip.
- Market heat: longs, shorts, or the combined tape — ranked, not alphabetized.
- Funnel counts: universe, shortlisted, scored, watch, armed, triggered, dropped.
- Loop health: scan / score / trigger workers and how stale the last pass is.
- Armed-now and on-watch candidates with the same card chrome as the live desk.
Public pages do not stream those counts. Inventing a delayed heatmap would be a lie in Google’s index.
Twelve components. One composite. Caps on every sleeve.
Hunter does not average a vibe. Each sleeve has a raw reading, a weight, and a weighted cap so one screaming input cannot dominate the card.
Shape is the model layout, not a live name. Members see the real breakdown per candidate.
A score is permission to watch. A chain is permission to print.
Two lanes. Structure (sweep, CHOCH, BOS, volume, RR to the next pool, context) and crowded (funding tax, 1h BOS, OB/FVG retest, OI unwind). Incomplete chains do not become tickets.
Structure lane
Sweep → CHOCH → BOS. A name does not print because heat is high.
- 1Liquidity sweep
- 2CHOCH · 5m
- 3BOS · 5m
- 4Volume on break
- 5RR ≥ 2 to next pool
- 6Context clear
Crowded lane
Funding tax + higher-timeframe break + positioning unwind.
- 1Crowded funding
- 2HTF BOS · 1h
- 3OB / FVG retest
- 4OI unwind
A+ / A / B. C does not exist as a card.
Grades are the Hunter language for setup quality. They sit next to the score so you can see conviction without opening a mystery long.
Highest conviction. Structure, liquidity, and trigger quality all clear. Rare by design.
Tradable. Score and chain both pass. This is the working grade of the book.
Eligible but not pretty. Printed only when the remaining steps still confirm.
Never a card. Heat without a chain is noise. Hunter drops it.
Radar, watch, signals, history, performance — then the bot toggle.
Radar
Heat, funnel, loop health, armed-now. The cockpit.
Hunter watch
Candidates in WATCH / ARMED with the live trigger strip.
Hunter signals
Cards printed only after the chain confirms. Own history.
Hunter report
R, quality grades, trigger lanes — not blended with the style desk.
Bot engines
Hunter has its own auto-trader toggle. It does not inherit scalp risk.
Lab (shadow)
A third book for research. Never wired to Telegram or the bot.
What Hunter is
- A 24/7 ranker of the venue’s linear perps.
- A scored shortlist with explicit sleeves and caps.
- A trigger chain that can say no after the score said yes.
- A paper book you can audit in R, with its own Telegram lane.
What Hunter is not
- Not a guaranteed win rate, and not financial advice.
- Not a reason to market-enter every amber cell on the heat map.
- Not the same engine as scalp / swing / position — those books stay independent.
- Not a substitute for invalidation. If the stop is gone, the idea is gone.
Hunter, in plain language
Is Hunter just the style desk with a new coat of paint?
No. The style desk (sig_v6) fires five independent profiles on names that already have a location and a confirm. Hunter is a second engine: it ranks the entire perpetual universe first, scores the top of it against structure and liquidity, and only then runs a trigger chain. Own Telegram channel, own paper book, own bot toggle.
Does a hot name automatically become a signal?
No. Heat is eligibility. A name enters Watch after the L2 score clears the bar, Arms when the chain gets past a liquidity sweep or HTF break, and prints a card only when the remaining steps confirm (volume, RR to the next pool, context, or the crowded-lane unwind).
Can I stack a Hunter card onto a Swing ticket on the same idea?
You can ignore the process. The product will not do it for you. Hunter and the style desk are independent books. That is the point.
Are Hunter cards financial advice?
No. Paper tickets, invalidation, and R multiples are tools. Leverage can liquidate you. Past paper results do not guarantee future results.
Run Hunter on the live venue
Radar, watch board, printed cards, history, and the Hunter bot toggle are membership surfaces. This page is the textbook. The desk is the lab.