Byreal, a decentralized exchange (DEX) incubated by Bybit and built for the Solanaecosystem, is excited to announce RealClaw, marking a shift toward financial systems designed for agentic trading. The new product allows individuals to participate in onchain finance by delegating trading, liquidity deployment, and portfolio management to autonomous agents.

With RealClaw, Byreal introduces a new model of participation in onchain finance, where users express intent through chat or voice commands while agents handle execution. Users can connect a wallet, select a strategy, and trade onchain without manually having to code, or perform swaps, configure liquidity positions, or monitor markets in real time. Instead, agents can run strategies continuously, execute across time zones, and keep capital active even while users are offline. The initial launch is by invitation only.

The set-up completely transforms trading DeFi trading experience. Users will be able to execute actions directly onchain by issuing instructions via Telegram, and receive real-time updates, portfolio summaries, and alerts via messaging platforms, while maintaining control over execution through confirmation flows.

RealClaw integrates non-custodial wallet infrastructure powered by Privy, allowing users to retain full control of their private keys and assets.

The launch builds on Byreal's early momentum in agent-native infrastructure. Prior to RealClaw, the team introduced Byreal Agent Skills, a set of execution primitives that allow agents to interact directly with onchain liquidity, including querying pools, analyzing yield, and opening or closing positions without manual input. Byreal Agent Skills surpassed 1,000 users within three weeks of launch.

"RealClaw brings our agent-native thesis to real users," said Emily Bao, Founder of Byreal. "Byreal leaves no one behind in the AI agent revolution by democratizing access to autonomous agents. Users can deploy sophisticated agent strategies without coding or infrastructure management under two minutes. That's what agent-native infrastructure looks like at scale."