Yahoo Finance quote, news, history, lookup, markets, gainers, losers, and
trending pages all arrive server rendered (the AAPL quote page carries the
closing price in about 456 tokens of view against 325,000 of raw HTML), so
a definition ships for all eight. The README grows the comparison people
actually ask for: the same stock price task run through Claude Code and
Codex as they ship versus with oc, one live session each, wrong answer and
all. Codex read raw HTML and reported a price that is not on the page;
WebFetch's digest was cheaper than oc this run and the README says so
plainly rather than hiding it. Also a banner, badges, and an llms.txt
pointer, so both people and models skimming the repo can tell what it is.
