The world of professional analytics is undergoing a tectonic shift. I have carefully studied recent experiments by independent developers and can confidently state: conventional notions about expensive terminals like Bloomberg are becoming obsolete before our eyes. The key news is that Anthropic's Claude Code tool is now capable of connecting within minutes to live financial data covering more than 17,000 stocks and cryptocurrencies, as well as earnings reports, balance sheets, and cash flow statements.

This is not a futuristic concept but a working scheme. One analyst known in niche circles under the pseudonym CyrilXBT demonstrated how, with just a single command and the MCP (Model Context Protocol) protocol, Claude Code can be integrated with an external financial data service. According to him, the entire setup process takes about 60 seconds and consists of three simple steps: adding the server, authorizing via OAuth, and starting to work with natural language queries.

What this means in practice

Forget about manually gathering data from dozens of websites or writing complex API requests. Now an analyst can simply ask Claude Code about Apple's current price-to-earnings ratio, Tesla's quarterly reports for the last four periods, or Bitcoin's performance over the year—and receive a structured answer based on current market data.

This is a direct challenge to an established business model. The cost of a Bloomberg terminal subscription reaches $24,000 per year, and the emergence of such a combination is unlikely to please specialists accustomed to such expenses. We are witnessing the democratization of access to information that was previously the privilege of large funds and corporations.

The era of autonomous agents: scaling without limits

Moreover, this trend extends far beyond simple data access. The same CyrilXBT revealed details of the workflow of Boris Cherny, one of the creators of Claude Code. Cherny uses a smartphone as his primary interface, managing 5–10 active sessions simultaneously. Each of them is capable of spawning hundreds or even thousands of auxiliary agents to solve large-scale tasks. Dozens of such loops run continuously in the background, with some of them persisting on the server even when the laptop is closed.

However, serious risks also lurk here. The analyst highlights four critical elements for safe scaling: a full-fledged instruction submission interface, isolated parallel execution, automatic result verification without human involvement, and strict limits on time, costs, and volumes. It is these mechanisms that turn overnight agent work from a risky venture into a manageable task.

My expert view: This is not just another tool update, but a fundamental paradigm shift. The combination of AI reasoning with access to real market data creates a "research edge" that accumulates daily. Traders and analysts who master this toolkit first will gain a sustainable competitive advantage comparable to using insider information—only on a completely legal basis.