Crypto news

An analyst under the pseudonym CyrilXBT has presented a working setup that fundamentally changes the rules of the game for financial professionals. With just a single command and roughly 60 seconds of time, he connected Claude Code to live market data—from quotes for more than 17,000 stocks and cryptocurrencies to earnings reports, balance sheets, and cash flow statements. This is a full-fledged replacement for Bloomberg Terminal capabilities, and that is not an exaggeration.

How it works: three steps to market analytics

The connection mechanics are simple and elegant. The first step—inside Claude Code, a command is executed to add the external service financial-datasets via HTTP transport. This command specifies the MCP server address and registers it in the tool. The second step—authorization: the user enters /mcp and completes the OAuth procedure in the browser, after which the connection status can be checked with a separate command. The third step—working with data in natural language.

CyrilXBT provides specific examples: a query about Apple's current price-to-earnings ratio and market capitalization, an analysis of Tesla's quarterly reports over the last four periods, or the year-long price dynamics of Bitcoin. All of this is processed directly, without manually gathering information from multiple sources.

The key point—the cost of the alternative. According to the analyst's estimates, a Bloomberg Terminal subscription costs professionals approximately $24,000 per year. The emergence of such a combination is unlikely to please those accustomed to paying for exclusive access to data.

Who wins from the new combination

A sustainable advantage will go to quants, analysts, and portfolio managers who learn to combine Claude Code's reasoning with access to live financial data. This is not a one-time advantage—it compounds daily. In essence, this is about a research edge that grows with continued use of the combination in everyday work.

None of the steps individually is complex. The value arises precisely from connecting an ordinary tool with a stream of up-to-date market data in a single interface. Here, Claude Code acts as the shell through which the user accesses a third-party data provider, eliminating the need for manual information gathering.

Scaling: thousands of agents overnight from a smartphone

CyrilXBT also broke down the workflow of Boris Cherny, one of the creators of Claude Code. Cherny uses a smartphone as the primary interface instead of a laptop and maintains between five and ten active sessions simultaneously. Each of these can spawn sub-agents—sometimes hundreds, sometimes thousands overnight on larger tasks. Dozens of loops run continuously in the background, with some of them persisting on the server side even when the laptop is closed.

For such a setup, four elements are critical. The first—launching and redirecting tasks from a smartphone: a full-fledged interface for sending instructions is needed, not just receiving notifications. The second—isolated parallel execution: each agent works in its own context and workspace; otherwise, running many processes simultaneously will produce chaos, not scale. The third, which is most often overlooked—verification without human involvement: each agent's result is checked against something real—another agent, a test suite, or an explicit rule—rather than the agent's own report of successful completion. The fourth—strict stop conditions and a spending cap: a process running for hours unattended requires an absolute limit on time, money, and volume that does not depend on whether a person notices a failure.

My take: Connecting Claude Code to market data is just the first step toward democratizing professional analytics. When such tools become widespread, the competitive advantage will shift from access to information to the ability to ask the right questions and build reliable verification systems. Those who master these skills today will gain a head start that will be difficult to overcome.