Algorithmic pricing tools are widespread in digital commerce, but how their design shapes market outcomes remains poorly understood. We study this in a controlled market experiment where participants design rule-based pricing algorithms using a dashboard interface and compete against each other over 50 periods in a sequential Bertrand game with a buy-box feature. We vary interface design across treatments: some participants receive pre-configured cooperative algorithms, others can consult a large language model for pricing advice. Our results show that most treatment variations lead to substantially higher market prices, driven by the start price and the rules of the algorithm. Pre-configured tit-for-tat algorithms are particularly effective at pushing prices upward, while LLM advice does not produce additional increases. Direct observation of algorithm choices confirms a shift toward more cooperative strategies across most treatments. These findings suggest that seemingly small design choices embedded in pricing dashboards can meaningfully facilitate supra-competitive outcomes with implications for competition policy and platform regulation.