Exchanges adding vibe coding draw builders, bots and volume

In an opinion piece, Ignacio Aguirre Franco, chief marketing officer at Bitget, wrote that exchanges that add natural-language “vibe coding” tools and open APIs are drawing independent builders, automated trading agents and increased trading volume. He cited the 11-person decentralized exchange Hyperliquid, which the piece says processed about $3 trillion in volume in 2025, as an example of the trend.

Aguirre Franco described vibe coding as a workflow in which a user states intent in plain language — for example, “set up yield across four chains” or “build a CEX/DEX arb bot” — and AI converts that intent into executable code. According to the piece, the generated code handles orderbook logic, gas optimization and risk calculations so developers and agents can deploy strategies faster.

The opinion notes that solo teams and small groups can use APIs and AI-assisted development tools to build trading terminals, bots and yield aggregators in days or weeks rather than months. It cites a case in which a scalping terminal called Hyperscalper was built in weeks using exchange APIs and AI tools. The piece states Hyperliquid’s ecosystem includes more than 180 active builders who have generated roughly $74 million in revenue.

Bitget’s contribution to the described model includes an Agent Hub that connects AI agents to trading infrastructure via APIs and software components called skills. The company also offers a command-line interface called bgc that the opinion says allows developers to call Bitget’s full API suite from shell commands with JSON output. The piece describes agents that detect trading intent from natural-language prompts and execute spot, futures and copy trades, and it references tools that enable large language models to call exchange functions directly.

The opinion addressed security concerns raised by critics about AI-generated smart contracts and unaudited bots introducing vulnerabilities. Aguirre Franco wrote that exchanges should require simulations, audit trails and pre-deployment reviews under conversational interfaces and provide institutional-grade risk controls. He added that speed should not replace safety and that AI-assisted workflows need transparency and traceability.

Aguirre Franco argued that exchanges that do not structure their data and APIs to support intent-driven, AI-assisted workflows risk losing volume to platforms that do, because builders and agents will route activity to venues where natural-language strategies are discoverable and executable. The piece also referenced industry projections for blockchain user growth and survey data indicating most new users do not routinely trade or earn yield, framing vibe coding as a response aimed at lowering technical barriers for those users.