Autonomous Operation

Once launched, the system runs itself. This is the launch's steady state and, in the intended case, the rest of its life.

The Treasury watches one thing — the canonical pool — and does two things. When the price sits at or below its accumulation anchor, it buys, at intensity scaled to the depth of the discount. When fresh base asset flows into the pool and the price shows real strength over the floor, it sells a bounded fraction of that inflow. Each liquidation's proceeds split in the same transaction: team share to vesting, supporter share to the claim pool, market-making share retained (and, on third-party launches, the ecosystem share to $PURE). All of it is condition-gated, capped, and observable block by block.

Execution stays pull-based throughout: permissionless entrypoints, an off-chain keeper making the routine calls on the sampling cadence, and no action ever taken unless the coded conditions hold at that moment. Nobody — including the team that launched it — holds a lever over any of this.

Around the autonomous core, ordinary market life proceeds. Anyone trades on the canonical market. Anyone may add liquidity independently. Holders vote if governance questions arise; supporters claim as their share accrues; the team claims vested compensation. Every one of those interactions runs through its own narrow contract surface, and none of them can reach into the engine.

Nothing about this state assumes growth. If demand fades, liquidation simply stops (no fresh inflow means nothing to sell from), accumulation continues only at value-justified prices, and capital sits preserved — passively, indefinitely. Decline triggers no emergency behaviour and no intervention; the system's response to a quiet market is quiet. What an actual ending looks like is the subject of the next page.