Mrmarket.ai

What mrmarket.ai does

The mrmarket.ai MCP server provides a research layer for US-listed stocks and ETFs. It translates analyst-style questions into structured results instead of requiring separate API calls for each filter or dataset. Supported work includes stock screens, rankings, company comparisons, historical series, cohort-relative analysis, and event studies.

The covered data includes daily adjusted prices, annual and quarterly financial statements, earnings actuals and estimates, insider transactions, valuation snapshots, and derived measures such as ROIC, free cash flow, margins, returns, debt-to-equity, and earnings surprise. The service covers more than 11,000 US-listed stocks and ETFs across NYSE, NASDAQ, and AMEX, with more than 20 years of financial history.

How it works

Describe the analysis in plain English through query_data. The engine resolves the requested companies, fields, periods, and relationships, then returns rows and columns together with the assumptions applied. Point-in-time handling is intended to prevent future information from entering historical or as-of studies. Unsupported or ambiguous requests produce explicit errors rather than inferred results.

The mrmarket.ai MCP server supports multi-dataset questions in one query, such as combining insider activity, cash-flow growth, and price conditions. Larger research tasks can be split across calls and orchestrated by the connected assistant. Responses include a stable query identifier, warnings, assumptions, and error codes for troubleshooting.

Setup and configuration

Connect to the remote MCP endpoint at https://mcp.mrmarket.ai/mcp. The connection uses OAuth during first use and is associated with a free account so that credits and rate limits can be applied. Account management and connection setup are available at mrmarket.ai/connect.

Claude Code can add the server with its HTTP transport command. Claude Desktop and ChatGPT use their custom connector settings, while Gemini adds it as a remote MCP extension. For Claude Desktop, use the connector interface rather than editing claude_desktop_config.json, because this service is remote rather than a local command-based server.

Tools and capabilities

• query_data runs metered analytical queries.
• fetch_page retrieves additional pages using the cursor returned by a large query; it does not run or charge for a new query.
• describe_data helps browse or search the available data catalog.
• get_symbols resolves tickers, sectors, industries, or the full supported universe.
• getting_started provides verified example prompts organized by use case.
• recent_queries exposes recent queries for context and reruns.
• get_account_status reports the account, plan, credits, and rate limits.
• report_issue submits a result or feature issue using a query ID.

Only query_data consumes credits. Its cost depends on query complexity; the other listed tools are free.

Limitations and notes

The mrmarket.ai MCP server is designed for research and analysis, not brokerage execution or real-time monitoring. Prices are end-of-day. It does not cover intraday or tick data, options chains, news or transcript text, raw filing text, macroeconomic series, foreign exchange, crypto, bonds, or futures.

Typical responses are described as arriving in under 10 seconds, but timing can vary by request. Rate limits and credit balances depend on the connected account. When a question is unclear, a field is unavailable, a cursor expires, or coverage is missing, inspect the response error code, warnings, and assumptions before retrying.