Uncle Stock MCP Server

The UncleStock MCP (Model Context Protocol) server lets AI assistants — including Claude, Cursor, and other MCP-compatible clients — query Uncle Stock's fundamental stock screener directly in conversation. Ask an AI to find undervalued stocks, compare quality scores, or screen by any of Uncle Stock's 1,800+ metrics without leaving your chat.

Server URL: https://www.unclestock.com/mcp
Transport: Streamable HTTP (JSON-RPC 2.0 over HTTP POST)
Version: 1.0.0 (last updated 2026-03-09). See the changelog for full history.
Quota: 5 free search_stocks calls/day — unlimited with Gold (find_metric and get_universes are always free). See pricing and subscribe.

Features

Tools

find_metric Search for metrics by keyword. Returns IDs, names, descriptions, scale data, and an isPercentage flag. Always call this first to discover metric IDs before filtering or sorting. Wraps the Indicators API.
search_stocks Screen stocks using filters, sorting, and extra metrics. Returns matching stocks with fundamental data. Accepts the same JSON structure as the Search API.
get_universes Get valid enum values for geography and industry filters (continents, countries, exchanges, sectors, industry groups, industries). Wraps the Universes API.

Authentication

Setup

Claude.ai (web)

  1. In Claude.ai, go to Settings → Connectors → Add custom connector
  2. Enter the server URL: https://www.unclestock.com/mcp
  3. Open the advanced / OAuth settings and enter:

    Log in to Uncle Stock first to see the Client ID and Client Secret.

  4. Save — Claude.ai will redirect you to the Uncle Stock login and consent screen
  5. Sign in with your Uncle Stock account and click Allow

Claude Desktop

Add the following to your claude_desktop_config.json (replace with your actual API key):

{
  "mcpServers": {
    "unclestock": {
      "url": "https://www.unclestock.com/mcp?apikey=your-api-key-here"
    }
  }
}

For MCP clients that support custom headers, you may also pass the key as X-Api-Key.

Usage examples

Example 1: Multi-region quality growth screen

"Looking for 8 best fitting stocks. Lets go for India and South America. I need fast and stable growers at a reasonable price. Look at cashflows. I need a healthy balance sheet so they have time in a period of downturn. For sectors I believe in healthcare, tech, infrastructure and consumer goods. Make sure the companies are valued acceptably."

The AI calls find_metric to look up revenue growth, free cash flow, balance sheet strength, and valuation metrics, then calls search_stocks filtering on India and South America, the requested sectors, and the financial criteria — returning 8 ranked results with key fundamentals.

Example 2: Multi-criteria safe stock screen with earnings integrity checks

"I am looking for 8 best fitting stocks. I need safe stocks this time. Have a look at sectors like infrastructure, construction, consumer goods. I want to invest in USA, Europe, England, Canada and Australia. I need stable good profits and a healthy balance sheet. Please check for earnings manipulation or insider issues. For Europe I need a good dividend, not for the other regions. Market capitalisation has to be between 200 million and 10 billion. Spread between sectors and regions."

The AI uses find_metric to identify the Beneish M-score (earnings manipulation), Piotroski score (financial health), and dividend yield metrics, then calls search_stocks with market cap bounds, multi-region geography, sector filters, and the integrity checks — spreading results across sectors and regions as instructed.

Example 3: Custom investment style screen

"I am looking for 8 stocks. Make a new score based on the investment style of the fund manager Greg Abel. Explain yourself. Make a score, define filter criteria on financial ratios, exclude regions and sectors."

The AI reasons about Greg Abel's known investment philosophy (capital allocation discipline, durable free cash flow, conservative balance sheets), calls find_metric to discover the relevant UncleStock metrics, and fetches the raw values via search_stocks. It then computes a custom weighted composite score client-side — combining Quality score, Piotroski score, Value score, and inverse Debt ratio — ranks the results, and explains the full methodology before presenting the top 8.

Example 4: Peer comparison

"Compare the Moury Construct with its peers. Look at cashflows, valuation and growth. Give me 8 similar stocks ranked from best to worst in quality."

The AI looks up Moury Construct to determine its sector and industry, calls get_universes to identify the correct industry filter, then calls search_stocks to find peers in the same industry filtered by cashflow, valuation, and growth criteria — returning 8 results ranked by Quality score with Moury Construct included for reference.

Quota

Privacy

The MCP server receives the queries you send via your AI client and returns stock data from the Uncle Stock database. No query content is stored beyond standard server logs. See the Uncle Stock Privacy Policy for full details.

Support

Questions or issues? Reach us at peter.nees@unclestock.com, or via the Uncle Stock in-app chat assistant. Problem reports and feature requests are welcome.