Market data MCP for research agents

Choose the narrowest market tool

Use symbol search when a company name is ambiguous, a quote for the latest available end-of-day observation, and price history for returns, volatility, or event-window calculations. Company profile answers identity and classification questions; fundamentals returns reported statements, ratios, and key metrics.

Earnings and analyst-ratings records answer different questions. Earnings data covers scheduled or reported company events and surprises. Analyst data summarizes third-party opinions and recent rating changes; it is not an Unquant recommendation.

Dates and data quality matter

The quote and price-history datasets contain daily bars and are unsuitable for live trading or order execution. Fundamentals should be labeled with their fiscal period. Earnings dates and estimates can change before a report, and ETF holdings are periodic snapshots that may lag portfolio changes.

A useful agent response states the observation date, reporting period, and known delay instead of describing every value as current. The public schema pages below list filters, row limits, primary keys, and field meanings for every market dataset.

Example research sequence

• Resolve each company to a normalized ticker.
• Retrieve the latest reported fundamentals and keep the fiscal periods.
• Fetch a bounded daily price window for market performance.
• Add earnings or analyst context only when it answers the user's question.
• Explain delays and avoid execution-grade claims.