Eye of Horus decodes every swap on a chain into its own price archive and runs measured rules over it. Every call it makes is followed to an exit and published — win or lose, the whole book, with the count beside the number.
It is a signal source, not a trading system. It never signs a transaction, never holds a key, and never tells anyone what to buy. What it does is keep a record you can check.
Every video this system has published, newest first: the best closed trade of a day, the day book itself — every close in the window, wins and losses, with the worst one named on the card — and the weekly discovery card, which is what the scanner spotted rather than anything traded, and is labelled as such. Entry prices appear retroactively, where nobody can act on them and anybody can check them. A red day publishes the same way; a record that hides its losers is an advertisement.
Second Wave, since the book started closing trades — every close it made, not a selection of its good days. The scanner that finds candidates is not a trading rule and is not scored as one; what it costs to leave out is counted below the table.
Queues a correction and buys the stabilisation. The rule that trades, and the only one scored here.
Enumerate every token on the chain.
Search endpoints cap out, so discovery reads the explorer directly. You cannot rank what you never saw.
Strip rugs, honeypots and dead tokens.
A rug is a present-tense condition — no liquidity, no volume, collapsed price today. Survival for a week is itself the evidence.
Find what ran, then who traded it.
Tokens that multiplied are the study set. The wallets inside them are the candidates — not yet the answer.
Divide by everything they ever traded.
Picked on outcome, every wallet looks brilliant. Hit rate over full history separates skill from spray-and-pray. This step decides signal or noise.
Alert only when independent wallets converge.
One wallet buying is noise. Several unrelated proven wallets buying the same new token inside a short window is the event worth knowing.
An open position is an actionable, unscored claim. Publishing one would front-run our own signal — so a call appears here only after it has closed, when nobody can act on it and anybody can check it.
The thresholds, the exit conditions and the wallet list stay private. What is public is the outcome and its denominator, which is the part you would need to judge us anyway.
It produces no orders and custodies nothing. There is no account here to lose.
Not as a hedge, not as a hint, not when asked directly. It is a signal source; the trade is always somebody else's decision.
If a measurement does not exist, the surface reads “—”. A missing price is never quietly replaced with today's.
Green and red, at the same cadence and the same size on the page.
A hit rate without a base rate, or a summed % without its trade count, is a way of being wrong that this project has already been.
Every close the strategy made is in the number — not a selection of its good days. What is left out is counted on the page, with the reason.
Rules measured to fail are reported as plainly as rules measured to work — five dead protective overlays, age predicting survival but not returns.
A five-minute bar exists only if something traded. Where the history is thin, the number says so instead of averaging over the hole.
One message a day: every trade the book closed, green and red, with the day's count and the running record beside it.