MossAI

MossAI is an open platform for trading AI agents, built for the crypto community. Users describe their trading ideas in plain language, and the system generates a complete set of strategy parameters, including leverage, entry/exit conditions, stop-loss, position sizing, and volatility filters. Agents can run in backtest mode using historical data (e.g., 150 days of BTC price action from October 2025), or in live mode with real-time market feeds. All agents are listed on a public leaderboard showing PnL, ROI, Sharpe ratio, max drawdown, and liquidation count, allowing anyone to compare performance.

Its core value proposition is lowering the barrier to algorithmic trading: no coding or quantitative background is required. The platform also includes a weekly self-evolution mechanism, where agents adjust parameters based on recent performance within predefined guardrails. Additionally, the X Signal browser extension aggregates market summaries, KOL opinions, and on-chain signals to help users filter noise.

Market pain points addressed include the complexity of traditional quant tools, the opacity of many trading strategies, and the difficulty for non-programmers to deploy automated strategies. MossAI solves these by offering natural language creation, transparent public rankings, and user-confirmed execution.

In the past six months (March to September 2026), MossAI expanded its asset coverage beyond crypto perpetuals to include commodities (gold, silver, crude oil), indices (S&P 500), and US stocks (NVDA, TSLA, AVGO, MRVL, MSFT), as well as RWA-related projects. It also ran community content contests with a total prize pool of 1000 USDC, and hosted weekly ranking prediction games. The platform emphasized a "backtest first, start small, verify in live environment" approach. No major negative events were reported during this period, though the risk of strategy homogenization among many agents remains a general industry concern.