feat(trading-signals-docs): move the trading knowledge to the website

[bennycode]

Domain knowledge lived in package READMEs, where it was hard to find and impossible to link to. It now has a Trading Basics section on the docs site, sitting between Getting Started and the indicator categories:

• Trading Glossary — the domain knowledge from the trading-strategies README, grouped into markets and infrastructure, orders and execution, positions and instruments, trader profiles, and evaluating a strategy. Kept as one page on purpose: a glossary is a lookup surface, so splitting it would turn every term lookup into a hunt across sub-pages.
• Indicator Types — the classification section from the trading-signals README, with the four function categories linked to their index pages and the named examples linked to their own.
• Risk Management — a new page covering why drawdowns are asymmetric, position sizing from a risk budget, stop placement (fixed, volatility-based, structural), fixed vs trailing stops, the win rate a given reward:risk actually needs, and the limit-vs-market trade-off on the way out.

Both READMEs keep a one-line summary that links to the page, so nothing is lost for people reading on npm or GitHub.

Closes #1289

Also in here

getPositionSize is added to the trading utilities. It answers "how many units do I buy so that being stopped out costs 1% of my account", which is the bridge between a stop price and a strategy's order size, and the one place where getting the arithmetic wrong costs real money.

Three decisions worth a look during review:

• It returns a fractional size rather than rounding. Flooring would turn 0.0431 BTC into 0. Callers round to the exchange's step size, which the glossary now explains.
• It uses the absolute distance between entry and stop, so shorts (stop above entry) work through the same call instead of returning a negative size.
• It throws rather than returning NaN or Infinity. The subtle case is NaN: every comparison against it is false, so a NaN input slipped past all four range checks and came back as a NaN position size. A single finiteness check on the result covers that, infinite inputs, and overflow from inputs that are each individually finite.

The risk management page deliberately carries no code — it is conceptual trading knowledge, and the library reference for the utility lives in the README table and on the Utility Functions page.