July 16, 2026

An AI Research Workflow: StockMarketScan MCP + EarningsCalls API

Combine the StockMarketScan MCP server with the EarningsCalls API to let an AI agent screen for setups and read the earnings call in one workflow.

Connect your AI client with the StockMarketScan MCP server.

The most useful stock research combines two very different data sources: structured market data that tells you what is moving, and unstructured earnings-call text that tells you why. Historically, stitching those together meant switching between tools and copying data by hand. With AI agents and the Model Context Protocol (MCP), you can hand both sources to the same assistant and let it do the joining for you.

Two tools, one agent

StockMarketScan exposes its screeners, chart patterns, options flow, and market data through an MCP server and a REST API, so an AI client like Claude Desktop, Cursor, or Continue can query screeners directly and get back structured results. Our sister product, EarningsCalls, provides earnings-call transcripts, summaries, and speaker data through its own API and MCP interface. Connect both to the same agent and you have covered both halves of the research problem: the quantitative screen and the qualitative narrative.

What the workflow looks like

A single natural-language request can now span both products. You might ask the agent to pull today's Hot Prospects screener from StockMarketScan, then, for the top names, retrieve the latest earnings-call summary from EarningsCalls and flag any where management raised guidance. The agent calls the StockMarketScan MCP tools for the screen, calls the EarningsCalls API for the transcripts, and synthesizes a shortlist with both the technical setup and the fundamental context attached. What used to be an afternoon of manual work becomes one prompt.

Why MCP makes this practical

The value of MCP is that it standardizes how tools describe themselves to an AI, so the agent can discover and call them without custom glue code. Because both StockMarketScan and EarningsCalls speak the same protocol, adding the earnings layer to your screening workflow is a matter of connecting a second server, not writing an integration. For developers who prefer direct access, both products also offer a straightforward REST API, so you can build the same pipeline into scripts, dashboards, or automated research jobs.

Getting started

Start by connecting the StockMarketScan MCP server to your AI client and confirming you can query a screener. Then add EarningsCalls as a second source for the transcript layer. From there, the workflow is only limited by the questions you think to ask. The screen finds the candidates; the call explains them; the agent ties the two together.

Related Resources