the EE is Exit Engine
Every tool is built for the entry.
Entry is a race you cannot win.
Median hold time on a Solana memecoin is about 100 seconds. Co-located bots beat you to the block by 400 milliseconds. Roughly 87% of same-block snipes are already green before you have even seen the ticker. You will not out-enter them, and no dashboard changes that.
The exit is different. It plays out over minutes, sometimes hours, and that is a timescale a person can actually work in. It is also where most people lose the trade: around half of pump.fun wallets finish a month down, and 96% end flat or worse. Most of them did not pick a bad coin. They rode the 40× back to 2× and then rode it the rest of the way to zero.
mEEme.xyz only does the second half of the trade.
69% on 156 exit calls graded · winners average +38%, losers -64% · every one public, wins and losses.
DOGEFI is being distributed right now — insiders have sold 67.8% of their bag.
- coil score
- 0.57
- confidence
- 100%
- first exit
- $0.01022
- hard stop
- $0.00687
Every holder has a cost basis
You are not the only one holding this.
Every unit of float was bought by someone, at some price. Reconstruct those prices and you know who is up, who is stuck, and what each of them does next.
what you actually get
Not a score out of ten. A call, the evidence behind it, and the exact prices to act on. This is a live run of the same engine that serves the app.
DOGEFI is being distributed right now — insiders have sold 67.8% of their bag.
- Insider cluster: 12 linked wallets hold 15.1% of tradable float at an average cost of $0.000480 — currently 21.3× up.
- That cluster has already sold 67.8% of what it held. Distribution is not a forecast here, it is in progress.
- Coiled supply 38.4% vs trapped supply 16.0% — more of the float is sitting in profit than is stuck underwater, so rallies get sold.
- Velocity of realization is +0.62 — profitable holders are actively converting to cash.
Take 51% at market now · 24% at $0.011303 · 16% at $0.017585 · 8% runs · hard stop $0.006866 (-33%).
- coiled
- 38.4%
- trapped
- 16.0%
- insider coil
- 15.1%
- trapdoor
- $0.00298
the mechanic
The candles do not show where a memecoin goes next. The unrealized PnL of everyone already holding it does.
Coiled supply
Every holder cheaper than you is a spring pointed at your exit, weighted by how far up they are and whether they have already started selling. Someone up 50× can dump on you any second. Someone up 1.1× has nothing to run from.
Trapped supply
Every holder more expensive than you is a bag that will not sell into weakness, because there is nothing to take. That is why a level 'holds' right where 8% of supply finally gets back to breakeven.
Insider coil
The same math, run only on wallets we can trace back to the deployer's own funding. Anyone can tell you insiders exist in a launch. We tell you what they paid, and exactly how much of it they have already sold.
what comes out
Not a score. A ladder and a trapdoor: the price where the largest block of in-profit supply goes to breakeven and paper gains turn into a stampede for the door. That price is your stop. It comes from the order book itself, not a round number someone picked because it looked clean.
why this is unfair
- It is non-consensus data
- RugCheck tells you a token is risky. DexScreener tells you the price. Neither one will tell you what the people about to dump on you actually paid. That number has to be derived, and nobody else is selling it.
- Latency does not kill you here
- An exit decision plays out over minutes, not milliseconds, which is a timescale a web app can actually compete in. Entry never had that luxury. Every entry tool is fighting a bot it cannot beat.
- It beats your own hands
- Memecoin edge is asymmetric: you win maybe 15 to 25% of the time, and make 3 to 10× on the ones that hit. A ladder set in advance holds a runner longer than you would and cuts a loser before you would.
Every call is public. Including the ones that were wrong.
Every tool in this space claims a win rate. Almost none of them will show you the rule they used to measure it. Ours sits in the repo, in plain code, and grades every call four hours after it fires with no human touching the outcome. A call that lands in the noise gets marked neutral and dropped from the average instead of being quietly counted as a win, and demo reads never make it onto the ledger at all.
That ledger is also why the engine stopped making entry calls: split by what a call told you to do, the exit side graded 72% and the entry side graded 21% — so the entry side was retired rather than tuned. The engine reads who still has to sell. It turned out to have nothing to say about who wants to buy, so it stopped saying it. See the live ledger for what it’s calling now.
Read the ledgerPoint it at a bag you already hold.
Three free locks a day. No account, no wallet connect, ladder included in the first read. If it gets the call wrong, that shows up on the ledger within the hour, same as everything else.
Open the cockpit