Vector 01 / Same-slot clustering
Four or more buys landing in a single slot is the cheapest label a classifier can apply. Volume is irrelevant — simultaneity is the tell.
Coordinated execution across thousands of Solana wallets. Every wallet arrives with its own history, its own funding, and its own slot.
Wallets managed 41,800
Waves run 12,406
Median spread 9.2 slots
01 — The problem
Cluster classifiers don't flag size. They flag shape — buys landing in the same slot, wallets born in the same hour, funding that traces back to one source. Three signals, and a campaign resolves into a single entity.
Vector 01 / Same-slot clustering
Four or more buys landing in a single slot is the cheapest label a classifier can apply. Volume is irrelevant — simultaneity is the tell.
Vector 02 / Birth synchrony
A wallet's first trade is a global timestamp. Wallets that start life together stay grouped forever, no matter how they behave afterwards.
Vector 03 / Funding graph
One source fanning out to a hundred wallets is a single graph edge away from collapse. The money path is the weakest link in most setups.
Fig. 01 — slot × wallet occupancy
Each column is one slot, each row a bucket of wallets. The cluster score is the share of all activity falling in the busiest column — this is the picture a classifier sees.
02 — Method
Drawn from a pool trading for weeks before you need it. Ages staggered by construction.
Breaks birth synchrony
Exchange legs and multi-hop proxy chains. Each money path terminates somewhere different.
Breaks funding graph
Buys spread across slots, sizes drawn from a distribution, decoys woven through the wave.
Breaks slot clustering
We run the public classifiers against your campaign before it is live and report what they see.
Verifies 01–03
Per-wallet PnL, fee harvesting, and an unwind that does not rebuild the graph you avoided.
Preserves funding graph
03 — Fit
Built for
Not for
If you are in the right-hand column, the first reply will say so.
04 — Operations
A campaign is only as quiet as its least careful component. Each stage publishes a health beacon, and execution gates on all of them.
Swipe for the rest →
05 — Record
Every campaign is scored against the same public cluster classifiers before it goes live and again after it settles. Those tools are open to anyone, including you — so this is a claim you can check rather than one you have to trust.
Public bundler index — lower is quieter
06 — Access
Tell us what you are launching, the size you need to move, and when. If we cannot carry it, the first reply will say so.
You pay when you withdraw, not when you trade.