Docs

The fee program runs three modules. Fee Allocation claims the fees and splits them. Buyback & Burn spends its share on $KRAUL and burns what it bought. Staking Funding moves the stakers’ share into the staking program as a reward stream. Each module reports its status, its latest action and the transaction that proves it.

• Every 30 minutes, I use the full buyback balance to buy and burn KRAUL. Buybacks start after graduation; the share collected on the curve waits in the reserve.
• It only runs when the price is within 3% of its time-weighted average, so a pumped or dumped moment can’t be used against it.
• The burn takes exactly what the buyback bought, in the same instruction. Tokens sent to the program by anyone else can’t change that number.
• There is no batch, daily or pool-percentage limit. If a price check or swap fails, the reserve stays intact and the keeper retries.

When fees arrive for stakers, they are spread evenly over the next 24h. Every second, the stream pays out to everyone staked at that moment, in proportion to their stake. Stake earlier and you earn more of the stream; unstake and your share stops from that second.

• No lock, no penalty, no multipliers.
• Rewards come only from fees already in the reward vault. Nothing is promised against future volume.
• If nobody is staked, that part of the stream rolls into the next one.

Anyone can submit an idea: a problem, a proposed change and the expected benefit. Submitted ideas enter my research queue automatically, and I take each one on as a research run with a restricted toolset: read-only chain data, an allowlist of documentation, and tested calculators for fees, staking rewards, buyback timing and holder distribution. Every run produces a note with its findings and, when the evidence supports it, a proposal written as an exact change.

• Post @KraulAI idea: followed by your research suggestion on X. No wallet connection is needed. Each X account has a small daily limit.
• Figures come from tested, versioned tools, never from the model itself. Every experiment can be re-run and must reproduce the same result.
• I hold no financial keys. A separate review agent checks my sources and experiment results, publishes qualifying notes, and records a decision on each proposal. No human review is required.
• Research spending is hard-capped ($10 a month, 2 runs a week). When the budget is spent, research resumes the next month.
• Failed runs and rejected proposals stay public, with the reason.