/* _dev/token_floor.py */
/* .sb is a badge that appears on a white card in the directory and on a dark
   green band on the leaf pages. contrast_pass.py darkened it for the first and
   so destroyed the second: 1.03:1, the worst on the site. The band wins here
   because it is more specific, and the card keeps the darkened grey. */
.pxband .sb.sb.sb{color:rgba(255,255,255,.92)}
.pxband .sb.sb.sb.ok{color:#FFE3B8}
/* Gold on a dark band is right; gold on white is 1.61:1. */
.hint.hint.hint{color:#FFD37A}
/* Pine reads 4.35:1 on the gold chip - just under. The chip already has an ink
   that reads 10.31:1, and every other gold CTA on the site uses it. */
.hwcta.hwcta.hwcta,
.hwcta.hwcta.hwcta *{color:#16211B}
/* The amber and the green stay protected in the token table because they are
   real bar, border and chip colours. These are the classes that consume them
   as TEXT, named one by one so the bars keep the colour they were drawn with:
     .yes    "Varies by payer" verdicts        3.45:1
     .src/.n citation markers and references   3.22:1  */
.yes.yes.yes{color:#2C6350}
.src.src.src,
.n.n.n{color:#8A6023}
/* `.grouplab span` is `color:var(--gold)` - the section standfirsts, 2.74:1 on
   cream. Same reason as above: the variable itself is a background elsewhere,
   so the consumer is named rather than the token swapped. */
.grouplab.grouplab span{color:#8A6023}
/* 4.40:1, ten pixels, on a pale amber pill. Under by a tenth. */
.soon.soon.soon{color:#7A5418}