QGold

What it is

An MCP server over a price history you host yourself. A scheduled Lambda pulls gold, silver, palladium, copper, bitcoin and ether prices from a public API each weekday evening and writes them to DynamoDB under a date index; the server reads that history back, so "what did copper do last month" is a date-range query rather than a single live quote. A second tool is unrelated to metals: it projects a starting principal over thirty years at a growth rate you set, withdrawing a percentage each year, optionally with random annual returns and with withdrawals skipped after a losing year.

What you get

• Prices for gold, silver, palladium, copper, bitcoin or ether by name, for a date range or for the most recent record
• A history that accumulates from the day you deploy, indexed by date rather than by timestamp
• A thirty-year projection table: starting balance, withdrawal, balance after withdrawal, growth rate, growth amount and ending balance for each year
• Random annual growth as an option, so the projection reads as a scenario rather than a straight line
• Withdrawals skipped in the year following a negative one, when you ask for that rule

Requirements

The server is the front end; the data lives in AWS resources you deploy yourself. That means an AWS account, the AWS and SAM command-line tools, and a deploy of the included template — two Lambda functions, a DynamoDB table, and an EventBridge rule that runs on weekdays at 22:00 UTC. Until that stack exists and has run at least once, the price tool has nothing to read. Python 3.10 or newer runs the server, and there is no key of the project's own to obtain.

Setup effort

Build from source — git clone https://github.com/jcaple/qgold.git

Tools you get

2 tools named in the project's documentation.

Requires

• An AWS account, with the AWS CLI installed and configured
• AWS SAM CLI
• Amazon Q Developer CLI, logged in
• Python 3.10 or later (as python3)

Where it runs

The documented client runs the server on your machine.