Market Prediction Address Bonereaper Faces Commission Doubts Over High-Frequency Short Cycles
A story circulating on social media claims that a Brazilian university student dropped out and created a market prediction trading bot using open-source agents in a garage, earning approximately $794,000 in fourteen months, with all code generated by Claude and monthly server costs around $10. The story points to the wallet named Bonereaper, which specializes in Bitcoin trading with five-minute fluctuations, employing Markov thresholds and automated parameter adjustments at midnight. This story has not been independently verified and is considered a replicable customer acquisition narrative.
On-chain data shows that the similarly named address is indeed a high-frequency participant: tracking pages show tens of thousands to millions of transactions, with transaction volumes reaching tens of billions, and realized profits and losses exceeding one million dollars, with a win rate of about 55% and a profit-loss ratio close to one. Community research indicates that a group of addresses, including this name, has enjoyed significantly higher commissions than publicly available rates since around August, with public rates peaking at about 50%, while accusations suggest rates up to 80%, later observed to have fallen to around 70%. The platform has not documented this arrangement as a public product description.
Nous Research's Hermes agent has indeed received large-scale open-source star ratings and has secured around $70 million in funding, including from Paradigm, with further discussions for higher valuations; this only proves the existence of the tool, not the truth of the dropout garage narrative. The five-minute trading window allows for programmatic price arbitrage and commissions, but is not suitable as a manual for "non-coders to replicate $794,000."
When the profit-loss ratio is close to one, profits are extremely sensitive to rates and commissions. If the commissions are true, the margins may come from the fee schedule rather than the myth of an 87% probability. If the commissions are false, real expectations still need to be reconstructed using public transactions and fees. Presenting the agent's midnight threshold adjustments as a holy grail obscures the fact that market making and incentive terms are the main variables in short-cycle trading.
Mechanically, the seller is the short-cycle event contract platform, while the buyer is a program capable of thousands of transactions daily. Event-driven profits are layered with commission accusations. Funds rapidly turnover within the five-minute window, with profits potentially coming from counterparties or undisclosed fee discounts. Beneficiaries are those who can access special commissions or delayed advantages; under pressure are retail traders operating at public rates and those treating the agent's narrative as a source of strategy.
Source: Public Information
ABAB AI Insight
The dropout garage and zero-code narrative serve as a customer acquisition template, while Bonereaper is a verifiable high-frequency address. The two are welded together to sell the illusion of an agent framework and a $10 monthly rental. The tracking page's win rate of 55% and profit-loss ratio of about one indicate that this is not an accurate prediction but rather an industrial flow with thin margins. If the 80% commission on trades is valid, the flow itself is a profit machine: the more transactions, the more returns, with Markov merely being the packaging.
The capital path is a fee war in the prediction market. Public rates exchange transaction volume for commissions, while private rates apply the same leverage to a single entity. Delayed trades are said to be shortened to make way for programmatic trading. Hermes and Claude are included in the tech stack to make it seem that the barrier is a 30-minute deployment, rather than settlement speed, inventory, and undisclosed terms. Paradigm's investment in open-source agents targets the agency layer, not guaranteeing the Sharpe ratio of this wallet.
Analogies can be drawn to forex commission networks, opaque exchange market-making agreements, and early high-frequency trading using custodial arrangements for fee reductions. The industry phase is a five-minute crypto trading environment turning prediction markets into micro-casinos, where incentive tables determine who can execute 40,000 trades daily. Those who only see the agent's replay miss the fee schedule as the strategy.
Structural judgments pertain to the transfer of pricing power. The mechanism is that the expected value of short-cycle contracts is rewritten by fees and commissions, while model probabilities serve as advertising. Once public rates and private discounts split, the same market does not represent the same game for two groups. Agents can adjust thresholds but cannot change the fact that they are not on the discount list.
ABAB News · Cognitive Law
A flow with a 50% win rate and a profit-loss ratio of one often has profits written in the fee schedule.
Zero-code wealth narratives sell tools, not replicable expectations.
Once undisclosed commissions appear, public markets are no longer the same game.