Robinhood post-selloff recovery-edge hunter. Reaper ranks sharp crypto selloffs by rebound participation, quoted spread, available depth, and recovery quality before a setup reaches the board.
Most assets trade close to fair after a selloff. Reaper looks for the smaller set where participation returns, spreads stay controlled, and the book still supports an exit.
SCAN -> MEASURE THE SELL-OFF -> CHECK RECOVERY -> SCORE -> HUNT OR SKIP
Snapshot: September 4, 2026 · 12:27 UTC
| Asset | Price | 24h move | Recovery score | Decision |
|---|---|---|---|---|
| LINK | $12.02 | +6.56% | 86/100 | HUNT |
| XRP | $1.45 | +5.84% | 78/100 | WATCH |
| DOGE | $0.08768 | +5.46% | 73/100 | WATCH |
| ETH | $2,521.04 | +4.75% | 68/100 | WATCH |
| BTC | $81,172 | +4.16% | 55/100 | SKIP |
| AAVE | $135.75 | +3.59% | 31/100 | EXPIRED |
The snapshot is fixed to the timestamp above. Reaper does not present historical values as a live feed.
Use case— rank recovery setups after a sharp selloffPrimary inputs— rebound participation, quoted spread, depth quality, and time since the lowPrimary failure mode— treating a bounce as proof that recovery is durableOutput— HUNT, WATCH, SKIP, or EXPIRED with the supporting score
Price alone does not prove that buyers returned. Reaper looks for participation that persists beyond the first reactive print.
A strong-looking rebound loses value when the quoted spread absorbs the edge. Reaper rejects candidates that remain expensive to enter or exit.
The board favors assets with enough visible depth to support the intended size. Thin setups stay out even when the headline move looks attractive.
Very early rebounds lack evidence. Very late rebounds often lose the asymmetry. Reaper scores the window between those extremes.
A candidate reaches HUNT only when all five checks agree:
- the quoted spread stays inside the configured ceiling
- recovery participation clears the minimum score
- the current move still sits inside the valid recovery window
- the book supports the intended size
- the setup is not a duplicate of an active name
Every rejected item keeps its reason. Quiet output means the board found no setup worth promoting.
REAPER // RECOVERY TICKET
asset LINK
price $12.02
24h move +6.56%
quoted spread 0.18%
depth deep
recovery score 86/100
decision HUNT
reason: participation returned while spread and depth stayed inside the operating limits
- pull the supported Robinhood crypto snapshot
- identify assets recovering from a meaningful session low
- score participation, spread, depth, and timing
- remove duplicate, late, or weak candidates
- rank the survivors and print one clear decision per asset
The loop stays deterministic around its gates. The model explains the setup after the data checks pass; it does not override a failed gate.
spread ceiling— blocks rebounds whose execution cost consumes the edgerecovery floor— requires enough participation before promotiontime window— removes rebounds that are too early or already staleposition cap— limits the number of promoted namesdry mode— keeps the scaffold observational unless execution is configured explicitly
git clone https://github.com/ReaperProtocol/Reaper
cd Reaper
npm install
cp .env.example .env
npm run devMIT
rank the recovery, not the drama.