// RULES · NOT A CONTROLLER
RULES
There is nothing to play. The engine is a decision stack. You can read every gate. You cannot type a fill into existence.
1 · Shadow
Default mode records the decision and a hypothetical mark. It does not sign, and this build has no key. Live routing is not in this build.
2 · Entry
The corrected candle replay signals on a completed close and enters at the next opening price, including fees and slippage. The baseline entry band is $3,000–$55,000. Research positions reserve 2% of equity, with at most three open at once and a 5% realized daily loss limit. These are starting research settings, not proven profitable parameters.
3 · Exit
The baseline tests a 12% stop, a 3× target, a trailing stop and a 15-minute holding limit. An adverse gap can lose more than the stop. Missing exits and unpriced migrations count as full-loss stress outcomes. A fixed USD market cap is never assumed to be an executable graduation fill.
Eighteen candidate rules are compared on older data. Validation includes higher slippage; the newest launches remain separate until selection is finished. The current data sample is insufficient to promote a candidate. Research status.
4 · Surplus
Only realized net, after fees. 20% to wallets holding strictly more than 1% of supply, pro-rata among that set. 20% buybacks of $LAND. 60% stays in the engine. No eligible wallet: the 20% joins the buyback. A loss pays nobody.
5 · What the agent is
LAND markets itself as pure intelligence aligned with the interests of capital. It is an accelerationist agent coin. It drafts. A human still releases the queue. It does not ask for a wallet.
Stress and paper results are not future returns.