原型 · 预览版：收录官方注册中心里的全部服务器，每 30 分钟同步一次。评价和账户功能将在后续阶段推出。
PopNx MCP为使用 AI 的人准备的 MCP 服务器 · 用大白话讲清楚
新手模式 开
配色跟随系统浅色深色
语言ไทยEnglish中文日本語한국어EspañolDeutsch
☰ 菜单设置
新手模式 开
配色跟随系统浅色深色
语言ไทยEnglish中文日本語한국어EspañolDeutsch

mrmarket.ai

作者：codinghaven版本 1.1.0i★ –i在云端运行
注册名称ai.mrmarket/mrmarket-mcp

安装

概览

作者提供的说明（英文原文）

Analytical engine for US-listed equities: screens, rankings, and event studies in plain English.

PopNx MCP 的三大支柱

这三项永远不会合并成一个分数：意见、测量和人气是不同的东西。

它能帮你做什么

我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。

安装

选择你使用的 AI 应用，然后复制设置。

安装之前

你需要

你用的是哪个应用？

在上方选择你的应用。我们会记住它，下次按钮会显示“安装到 …”。

适用于网页版、桌面版和手机版 Claude：

  1. 打开 Settings → Connectors
  2. 点击 Add custom connector
  3. 粘贴下面的地址，然后点击 Add
  4. 如果要求登录，请登录作者的服务
SettingsConnectors1+ Add2https://…3
https://mcp.mrmarket.ai/mcp

在终端中运行这条命令

claude mcp add --transport http mrmarket https://mcp.mrmarket.ai/mcp

在 Cursor 中打开

或者自己复制设置

粘贴到项目中的 .cursor/mcp.json，或 ~/.cursor/mcp.json

{
  "mcpServers": {
    "mrmarket": {
      "url": "https://mcp.mrmarket.ai/mcp"
    }
  }
}

浏览器会先询问是否打开应用，然后应用会显示设置并请你确认。密钥的值会留空，由你自己填写。

在 VS Code 中打开

或者自己复制设置

粘贴到项目中的 .vscode/mcp.json

{
  "servers": {
    "mrmarket": {
      "type": "http",
      "url": "https://mcp.mrmarket.ai/mcp"
    }
  }
}

浏览器会先询问是否打开应用，然后应用会显示设置并请你确认。密钥的值会留空，由你自己填写。

ChatGPT 只能连接互联网上的服务器：

  1. 打开网页版 ChatGPT → Settings → Connectors (Apps)
  2. 点击 Create 或 Add（部分套餐需要先在 Advanced 中开启 Developer mode）
  3. 粘贴下面的地址，起个名字，然后保存
  4. 如果要求登录，请登录作者的服务
SettingsConnectors1+ Create2https://…3
https://mcp.mrmarket.ai/mcp

菜单名称可能因套餐和版本不同而略有差异。

你的选择通过 Cookie 保存在这台设备上，无需账户。

这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同，请以 README 为准。

它能做什么

我们还没有读取这个服务器的工具列表；我们会轮流检查各个端点。

基础检查i

我们检查的是任何人都能看到的信息；我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月7日

  • ✓ 通过

    作者与仓库一致

    注册名称已通过域名 mrmarket.ai 验证（注册中心会核实域名所有权）

  • … 尚未检查

    仍在维护

    我们还没有从 GitHub 读取这个仓库

  • … 尚未检查

    有许可证

    我们还没有从 GitHub 读取这个仓库

  • … 尚未检查

    端点有响应

    尚未检查；我们会轮流检查各个端点

  • … 尚未检查

    使用 OAuth 登录

    尚未检查；我们会轮流检查各个端点

  • … 尚未检查

    工具描述中没有隐藏指令

    尚未检查；我们会在检查端点时读取工具列表

与“其他”中的其他服务器比较i

  • 星标数多于此分类中其他 18336 个服务器的 0%
  • 发布到注册中心的时间比此分类中其他 18336 个服务器的 10% 更新
  • 基础检查结果等于或优于此分类中其他 18336 个服务器的 100%

与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时，星标数才计入。

常见问题

mrmarket.ai 安全吗？

它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前，请先看看这个服务器要求哪些权限，并使用只具备所需权限的密钥。

它需要哪些密钥或代码？

注册中心没有列出必需的密钥，不过有些服务器会在首次使用时要求你登录。

它在我的电脑上运行还是在云端？

它在作者的云端运行。通过互联网连接，无需安装，但你发送的内容会经过作者的服务器。

它还在维护吗？

GitHub 仓库最后更新于–，最新版本（1.1.0）于5个月前进入注册中心。

哪些应用可以使用它？

支持 MCP 的应用，例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接，所以接受远程 MCP 服务器的应用也可以使用，例如开启开发者模式的 ChatGPT。

有什么可以替代它？

同一分类中有：Spec Driven Development, Remote Desktop Commander, NexusWindows

替代选择

同一分类中其他作者的服务器，按星标数排序。

如何看卡片：★ GitHub 星标i三大支柱：评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i

Spec Driven Development

作者：formulahendry

MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.

在你的电脑上运行★ 440MIT14个月前更新
安装

Remote Desktop Commander

作者：desktop-commander

Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.

在云端运行★ 211前天更新
安装

NexusWindows

作者：pironjulien

Windows computer use in the background: desktop apps driven without your mouse or keyboard.

在你的电脑上运行今天更新
安装

作者的 README

打开 README

按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。

mrmarket.ai MCP server

A purpose-built analytical engine for financial research, exposed over MCP. Ask hard questions about US-listed equities in plain English and get structured, reproducible results back.

mrmarket.ai is a remote MCP server. Connect it to Claude, ChatGPT, or Gemini and your assistant gains a dedicated engine for resolving financial questions against clean, structured market data. It is not a language model with a database bolted on, and it is not an API wrapper. You describe the analysis; the engine computes it and returns rows, columns, and the assumptions it applied.

The interesting questions in equity research require joins that no single API endpoint will run for you. "Stocks where insiders bought over $1M, free cash flow grew four quarters running, and the price sits below the 200-day moving average" spans insider filings, quarterly cash flow statements, and daily prices at once. mrmarket.ai is built for exactly that class of question.

Why it is different

  • Computed, not predicted. Answers are derived from the underlying data, not generated from a model's recollection. The same question returns the same answer every time.
  • Cross-dataset by design. Fundamentals, prices, earnings, and insider activity are resolved together in a single question, with time dimensions aligned across differing data frequencies.
  • Point-in-time aware. Historical and as-of questions respect the information available on the date in question, so backward-looking studies do not leak future data.
  • Built to fail loud. When a question falls outside coverage, the engine returns an explicit error rather than fabricating a plausible number.
  • Transparent. Every result carries the assumptions and defaults that were applied, so they can be inspected and adjusted.

Connect

mrmarket.ai is a remote MCP server reachable at:

https://mcp.mrmarket.ai/mcp

Each connection is tied to a free account so that credit balance and rate limits apply correctly. Authorization happens over OAuth on first use. Create an account and manage connections at mrmarket.ai/connect.

Claude Code
claude mcp add --transport http mrmarket https://mcp.mrmarket.ai/mcp
Claude Desktop

Open Settings, go to Connectors, choose Add custom connector, and paste the endpoint:

https://mcp.mrmarket.ai/mcp

Claude prompts for authorization on first use. Note that editing claude_desktop_config.json by hand does not work for this server; that file only accepts local, command-based servers, and mrmarket.ai is remote.

ChatGPT

In Settings, open Connectors, choose Add custom connector, and point it at the same endpoint.

Gemini

Add it as a remote MCP extension when prompted by Gemini's MCP onboarding, using the same endpoint.

Tools

ToolCostPurpose
query_datameteredThe workhorse. A natural-language financial question in; structured rows and columns out. Screens, rankings, comparisons, time-series, cohort studies, and event studies are all single calls.
fetch_pagefreeThe continuation. When a query_data result is too large for one response, page 1 comes back with a pagination.next_cursor; pass that cursor to fetch_page to retrieve the rest. It runs no new query, so it never costs credits — the original query_data is charged once.
describe_datafreeThe data catalog. Browse categories and fields, or search for a specific metric, before composing a question.
get_symbolsfreeFast ticker, sector, and industry resolution. Name to ticker, sector membership, or the full universe in under 50ms.
getting_startedfreeA structured orientation tour with verified example prompts, grouped by use case.
recent_queriesfreeYour recent queries, for context and re-runs.
get_account_statusfreeConnected account, plan tier, credit balance, and rate limits. A useful first call in any new session.
report_issuefreeFlag a result that looks wrong or a feature that is missing. Pass the query_id from the response so the trace can be investigated.

Seven of the eight tools cost zero credits. Only query_data consumes credits, priced by query complexity. Start with getting_started or describe_data, then move to query_data.

What you can ask

Every prompt below resolves in a single query_data call. Phrase questions the way an analyst would speak them, not as a list of column filters.

Screens
Technology stocks with ROIC above 20%, debt-to-equity below 0.5, and 5-year revenue CAGR above 10%.

Healthcare companies with positive free cash flow and gross margin above 50%. Companies with four consecutive quarters of growing free cash flow.

Rankings
Top 20 stocks by ROIC, excluding financials.

Top 15 companies by smoothness of year-over-year revenue growth over the last five fiscal years. Bottom 10 consumer discretionary stocks by 1-year total return.

Comparisons
Compare AAPL, MSFT, and GOOGL on revenue and net income over the last five years.

AMZN versus WMT operating margin trend over the last decade. NVDA quarterly EPS surprise history.

Cohort-relative analysis
Stocks whose ROIC is at least one standard deviation above their sector average.

Companies with a net profit margin more than twice their sector average. Companies with revenue growth in the top quartile of their industry.

Time-series and overlays
AAPL daily closes for the last five years with earnings dates overlaid.

MSFT price with insider buy and sell markers over the past two years. 50-day and 200-day moving averages for TSLA.

Event studies
Average 30-day return after companies beat earnings by more than 10 percent.

Forward 3-month returns following insider purchases above $1M, grouped by sector. Across the large-cap universe, the correlation between monthly insider purchase count and the next 3-month return.

Multi-step research, driven by your assistant

Larger studies decompose into several calls that your assistant orchestrates while mrmarket.ai supplies the ground truth at each step:

For every large-cap stock that gapped down 5 percent or more after earnings in the

past two years, what is the 30, 60, and 90-day recovery rate, and is it statistically significant against random 90-day windows?

Backtest this signal: buy stocks with insider purchases above $500K within 30 days before earnings that then beat EPS estimates, hold for 63 trading days, and report CAGR, maximum drawdown, and win rate against buying every post-beat stock.

Build a pre-earnings watchlist: companies reporting in the next two weeks where insiders bought in the last 90 days, the stock trades below its 200-day moving average, and they beat estimates last quarter.

Coverage

Universe11,000+ US-listed stocks and ETFs (NYSE, NASDAQ, AMEX)
History20+ years of financials on US equities
Response timeUnder 10 seconds for typical questions

Data spans daily adjusted prices, annual and quarterly financial statements (income statement, balance sheet, cash flow), earnings actuals against estimates, insider transactions, and valuation snapshots, alongside computed metrics such as ROIC, free cash flow, debt-to-equity, ROE, margins, returns, and earnings surprise. Sourced with point-in-time accuracy and curated for cross-dataset queries.

Scope

mrmarket.ai is a research and analysis layer, not a brokerage and not a real-time feed. Prices are end-of-day. The following are out of scope: intraday and tick data, options chains, news and transcript text, raw filing text, macro series, FX, crypto, bonds, and futures. When a question reaches an edge, the engine reports it rather than guessing.

Troubleshooting

Every query_data response carries a stable error_code, plus warnings and assumptions arrays that explain how the question was interpreted. Read those first; they usually point at the cause.

SymptomWhat it meansWhat to do
QUERY_NOT_UNDERSTOODThe question was ambiguous or referenced an unsupported metric.Rephrase with specifics (period type, time range, ranking criterion). Use describe_data to confirm a field exists.
RATE_LIMITEDYou hit your tier's per-minute or per-day cap.Wait briefly, or check get_account_status for your exact limits.
INSUFFICIENT_CREDITSYour credit balance is exhausted.Top up or upgrade at mrmarket.ai. Not a fault, no retry needed.
SCHEMA_DRIFTA data field is temporarily unavailable (mid-update).Rephrase without the affected scope and try again.
INVALID_CURSOR / SNAPSHOT_EXPIREDA fetch_page cursor was malformed or its result snapshot aged out (snapshots live ~15 minutes).Re-run the original query_data question to get a fresh result and cursor.
BILLING_UNAVAILABLE / SERVICE_ERRORA transient server-side issue.Retry once; the retryable flag will be set.
Empty result on a screenFilters were too tight.Loosen one threshold (e.g. ROIC > 15% instead of 25%) or widen the time range.
A result looks wrongAn assumption or default may not match your intent.Read the assumptions array. If it still looks wrong, file it with report_issue (include the query_id).

For a guided tour of capabilities, limits, and worked examples at any time, call the free getting_started tool.

Support

  • Report a wrong result, bug, or feature request with the report_issue tool. Pass the query_id from the response so the exact run can be investigated. This is the fastest path because it links directly to the query trace.
  • Email: support@mrmarket.ai

Links

问 PopNx