/*
	THE BUY RAIL -- THE TWO DELIVERY FACTS SIT SIDE BY SIDE (2026-11-19)
	====================================================================

	Operator, with the rims product on screen: "please dont touch the left
	side .. now i want change the right side, cause its looks like to
	messy. do you have another suggestion design?" -- then, from four
	treatments built on the mirror of the real page: **"B"**.

	WHAT B IS, OF THE FOUR. It is the one that KEEPS THE FULL WIDTH and
	uses it, rather than pulling the rail's content in to a narrower
	measure. The other three all narrowed it; this one spends the width on
	putting the two delivery facts beside each other instead of stacking
	them as two full-width bars.

	**SO THE 79-CHARACTER MEASURE IS A KNOWN, CHOSEN TRADE.** Measured at
	1430: the rail carries 630px of content where yucca -- the reference
	the 50/50 split was built to on 2026-11-18 -- carries 483, which puts
	the body text at 79 characters against a comfortable band of 55-75.
	Variants A, C and D each fixed that by capping the content at 420-460px
	and were not chosen. If "messy" ever comes back, that cap is the thing
	to reach for, and it is one declaration:

	    .sf-pdp__rail > * { max-width: 460px; }

	THE PHOTOGRAPH IS NOT THE PROBLEM AND WAS NOT TOUCHED. It measures 594
	square at 1430 against the reference's 603 -- the same size. Every rule
	in this file is scoped inside `.sf-pdp__rail`.

	WHY ITS OWN FILE. It is the house pattern for this page (pdp-gallery,
	pdp-options, pdp-related, pdp-reviews) and it keeps `single-product.css`
	-- 166KB that both the operator and this repo edit -- out of a change
	this small. Removing `pdp-rail` from the array in
	`superfly_pdp_components_assets()` restores the stacked rows exactly.

	The lab that produced it: `node docs/home-concepts/rail-lab.js`, then
	`docs/home-concepts/pdp-live/rail-lab.html`.
*/

/* ------------------------------------------------------------------ */
/*  Two facts, one row                                                 */
/* ------------------------------------------------------------------ */
/*
	`auto-fit` RATHER THAN A BREAKPOINT, and that is not laziness. The rail
	is 630px beside the gallery, about 700 on a tablet where the columns
	have stacked, and about 335 on a phone -- three different widths that
	do not line up with any breakpoint this page already has. `minmax`
	gives two columns wherever roughly 490px exist and one where they do
	not, so the phone keeps the stacked rows it has today with nothing to
	keep in step.

	240px is the floor because the longer of the two sentences -- the West
	/ East Malaysia delivery line -- needs three lines below about that,
	and three lines of 13px text makes the row taller than the pair of
	stacked bars it replaced, which is the whole saving gone.
*/
.sf-pdp__rail .sf-pdp__assure {
	grid-template-columns: repeat(auto-fit, minmax(240px, 1fr));
}

/*
	THE ICON GOES TO THE TOP, AND B IS WHAT MAKES THIS NECESSARY.

	`single-product.css` has carried `align-items: center` on this row
	since it was built, with a note recording the fault it causes and the
	fix: "with the text on two lines, `align-items: center` puts the icon
	BETWEEN them -- measured 9px below the first line centre.
	`flex-start` fixes it without changing anything else. Offered, not
	taken."

	It was fair not to take it then, because at full width each row was
	ONE line and the fault could not appear. Half width is two lines on
	both rows every time, so it fires on every product -- it is part of
	this change rather than a tidy-up smuggled in beside it.

	The 1px of top padding is optical: a 20px icon beside an 18px line box
	sits a hair high when both are flush to the top.
*/
.sf-pdp__rail .sf-pdp__assure-row {
	align-items: flex-start;
}

.sf-pdp__rail .sf-pdp__assure-row svg {
	margin-top: 1px;
}

/* ------------------------------------------------------------------ */
/*  The row has a measure again                                        */
/* ------------------------------------------------------------------ */
/*
	Operator, with the rims page open on a ~1890px screen: "but as you can
	see this, the size is too width .. and its not perfect, so how?" --
	then, from four treatments measured on the mirror: "Yes, cap the row".

	**THEY WERE RIGHT ABOUT THE PHOTOGRAPH AND I WAS WRONG.** Asked
	earlier the same day whether the picture was too big, I measured at
	1430, found 594 against the reference's 603 and said no. At their own
	width it is 809, and the rail's body text runs 105 characters a line
	against a comfortable 55-75. **Measure a complaint at the reporter's
	viewport before answering it.**

	THE CAUSE IS NOT THIS PAGE. `--sf-measure` went to `none` on
	2026-11-17 on the operator's own instruction, so the product row is
	full bleed and BOTH columns grow with the monitor without limit. The
	rail treatment chosen an hour before this was picked while looking at
	1430, where the same page measures 594 and 79.

	WHAT WAS CHOSEN, of four options put on localhost: row capped to 1400
	and centred, with the photograph taking 57% rather than 50. The slack
	then falls OUTSIDE both columns, where it reads as page margin -- a
	cap on the rail's text alone leaves it INSIDE the rail as a 335px hole
	down the right of the page.

	Measured at 1890 before and after: photograph 809 -> 651 (the
	reference is 603), text 105 -> 64 characters, the two delivery facts
	still side by side, 10px of slack beside the text.

	1400 IS SWEPT, NOT CHOSEN. At 1300 the rail's content drops below two
	240px columns and the facts fall back to stacked, undoing the
	treatment shipped an hour earlier. 1300/1360/1400/1440/1480 at 55-57%
	were all measured; 1400 at 57/43 is the first that keeps the
	photograph near the reference AND the facts beside each other.
*/
@media (min-width: 1001px) {

	/*
		`justify-content`, NOT `margin-inline: auto` -- AND THAT IS NOT A
		STYLE CHOICE. `.sf-pdp` carries a NEGATIVE inline margin (-101px a
		side at 1890) which is how it escapes its constrained parent to go
		full bleed. Setting auto margins to centre a capped row throws that
		escape away. The row is already `display: flex`, so centring its
		two children inside a still-full-bleed row costs nothing and leaves
		the escape alone.
	*/
	.sf-pdp {
		justify-content: center;
	}

	/*
		THE SPLIT IS SET ON THE RAIL ONLY, because only the rail carries
		one. Measured on the live page: the gallery computes `flex: 1 1 0`
		with NO inline style, and the rail carries `style="flex-basis:50%"`
		from its `wp:column` block -- so the gallery simply takes whatever
		is left and needs a basis here only to stop it doing that.

		`!important` is earned twice over: an inline style beats every
		rule in a stylesheet, and `templates/single-product.html` still
		says 50%. **If that template is ever edited to 57/43, delete these
		two declarations rather than leaving both.** CLAUDE.md records the
		template as the thing that decides the split; this file is
		deliberately overriding it so the whole change stays removable as
		one component.

		**A PERCENTAGE BASIS WITH A PIXEL CAP, AND A PIXEL BASIS IS A BUG.**
		`.sf-pdp` is a `wp-block-columns`, which computes
		`flex-wrap: wrap`. Fixed bases of 738 + 556 = 1294 do not fit until
		about 1400 of viewport -- and a flex row that cannot fit WRAPS
		before it shrinks, whatever the shrink factor says. Measured at
		1090 with pixel bases: the rail wrapped to a second line at y821
		under a gallery at y85, so the page stacked from 1000 to 1400 where
		it used to sit side by side -- and `html { overflow-x: clip }` meant
		the document reported no overflow at all while it happened.

		The percentage always sums to 100 and so can never wrap; the cap is
		what stops the columns growing past the measure on a wide screen.
		The hypothetical main size a flex container wraps on is clamped by
		`max-width`, so the cap costs nothing at the narrow end.
	*/
	/*
		900, NOT 738 (2026-11-19, the same day). Operator, shown the page
		capped: "but do you see that is it perfect interface UI? cause its
		looks like too many paddingg" -- and then, offered either one measure
		for the whole page or a smaller margin: **"Shrink the margin
		instead"**.

		**THE MARGIN ONLY SHRINKS IF THE PHOTOGRAPH GROWS**, and that is
		arithmetic, not a preference. The margin is (viewport - row) / 2 and
		the row is these two columns; the rail cannot grow because its cap IS
		the text measure, and widening it only moves the empty space inside
		the rail. So the gallery is the only lever.

		Measured at 1900, with the text held at 65 characters throughout:

		    cap 738 -> photo 682, margins 302   (32% of the screen empty)
		    cap 820 -> photo 764, margins 262   (28%)
		    cap 900 -> photo 844, margins 222   (23%)   <- here
		    cap 1000+ -> photo 936, margins 176 (19%)   <- the floor

		**900 IS A COMPROMISE AND IT IS WORTH KNOWING WHICH WAY IT LEANS.**
		The floor would take another 46px off each side, at a photograph of
		936 -- and the operator called 809 too big earlier the same day, so
		going there would ignore something they had already said. 844 sits
		just above the width the page had before any of this and halves the
		excess margin against that floor.

		IT ONLY BITES ABOVE ABOUT 1620. Below that the 57% basis is the
		smaller term and the cap does nothing: 1430 measures photo 685 and a
		61px margin either way, 1090 measures 497 and the page gutter.

		The dial that produced these: `node docs/home-concepts/row-width-dial.js`,
		then `docs/home-concepts/pdp-live/row-width-dial.html`. It prints the
		photograph, the measure, the margins and the empty percentage live and
		hands back the exact declaration, so moving this again is a drag.

		THE OTHER ANSWER, NOT TAKEN: the page carries TWO measures -- this row
		at 1294 and the reviews and related bands below it at 1740, measured
		at 1900. Capping those to match puts one edge down the whole page and
		fixes the same complaint without touching the photograph at all. It
		was offered and "shrink the margin" was chosen instead; it is still
		the better fix if this comes back.
	*/
	.sf-pdp__gallery {
		flex: 0 1 calc(57% - var(--sf-pdp-gap, 0px) * 0.57) !important;
		max-width: 900px !important;
	}

	.sf-pdp__rail {
		flex: 0 1 calc(43% - var(--sf-pdp-gap, 0px) * 0.43) !important;
		max-width: 556px !important;
	}

	/*
		AND THE TEXT STILL NEEDS ITS OWN CAP. The rail column is 556 here
		and its content 532, which is already close -- but on a product
		with a short title the blocks would otherwise run the full column
		while the body text wants the same measure as everything else.
	*/
	.sf-pdp__rail > * {
		max-width: 520px;
	}
}

/* ------------------------------------------------------------------ */
/*  Air at the top, and a real gap between the columns                 */
/* ------------------------------------------------------------------ */
/*
	Operator, 2026-11-19: "added margin top.. and i want you to added gap
	between section left and section right".

	WHAT WAS THERE. Measured live at 1900, scrolled to the top:

	    navbar pill paints to      y90
	    .sf-pdp starts at          y85   <- BEHIND the bar
	    photograph                 y110  (20px of air)
	    rail title                 y95   (5px of air)

	So the row began five pixels UNDER the floating navbar and the title
	all but touched it. The photograph looked fine only because
	`pdp-gallery.css` gives it its own 20px through `--sf-gal-pin-air`;
	nothing was doing the same for the rail.

	**THE HEADER IS STICKY AND IN FLOW, AND THE PILL HANGS PAST IT.** The
	header reserves 66px and the bar paints to 90 -- 24px of overhang that
	no following element knows about. CLAUDE.md records three other
	sections caught by exactly this. The padding below covers it and then
	some, rather than this file re-deriving the bar's geometry a fourth
	time.

	BOTH COLUMNS GET THE AIR, not just the rail. Padding the rail alone
	would start the title level with the photograph's second inch, and the
	two columns are read as one band.
*/
@media (min-width: 1001px) {

	.sf-pdp {
		/*
			THE GAP IS A TOKEN BECAUSE THE COLUMN BASES HAVE TO PAY FOR IT.
			The two percentages sum to 100 of the content box, so adding a gap
			pushes the row past what it has -- and a wrapping flex row wraps
			rather than shrinks. Measured at 1420 with a bare `column-gap`:
			741 + 24 + 556 = 1321 against 1300 available, and the rail dropped
			773px down the page. The SAME fault as the pixel bases before it,
			from a different cause, and `html { overflow-x: clip }` reported
			nothing wrong either time.

			Each basis gives up its own share of the gap -- .57 and .43 -- so
			the three terms come to exactly 100% at every width.
		*/
		--sf-pdp-gap: clamp(16px, 1.7vw, 32px);
		padding-top: clamp(20px, 2vw, 32px);
	}

	/*
		A REAL `column-gap`, NOT MORE PADDING ON THE COLUMNS. The 56px that
		was there is `.sf-pdp__gallery`'s 46px right padding plus the rail's
		10px left -- and the gallery is the left side, which this file is
		not supposed to touch. A gap on the ROW adds between the two without
		either column's box changing.

		**IT COMES OUT OF THE MARGIN, NOT OUT OF THE COLUMNS.** Both columns
		are capped, so the row simply gets wider by the gap and the side
		margins shrink by half of it each -- which is the direction the
		margin was asked to move in an hour earlier anyway.

		`column-gap` and not `gap`: below 1001px the columns wrap onto
		separate flex lines, where `row-gap` would govern and this must not
		add anything. The phone's vertical rhythm is left exactly as it is.
	*/
	/*
		AND THE SELECTOR NEEDS THE BLOCK CLASS. single-product.css sets
		`gap: 0` on `.sf-pdp.wp-block-columns` -- (0,2,0) -- which beats a
		lone `.sf-pdp`, and WordPress emits its own
		`.wp-container-core-columns-is-layout-…{ column-gap: 0 }` inline on
		top of that. Measured with one class: the rule was present, matching
		and losing, with the column gap still computing 0px.
	*/
	.sf-pdp.wp-block-columns {
		column-gap: var(--sf-pdp-gap);
	}
}

/* ------------------------------------------------------------------ */
/*  The name leads, and the review link stops shouting                 */
/* ------------------------------------------------------------------ */
/*
	Operator, 2026-11-19: "do you think the interface is perfect now? the
	font style the position and so on? if not pleaseee change yaa".

	**THE PRICE WAS BIGGER THAN THE PRODUCT NAME.** Measured on the live
	rail: title Archivo Black 28px, price Chakra Petch 32px. On a product
	page the name is the first thing being answered -- "what is this?" --
	and the price answers the second. A rail that sets the price larger
	reads as a price list with a picture beside it.

	THE TITLE GOES TO 36 AND NOT PAST IT, because 36 is the last size that
	still sets this product's name in THREE lines at the rail's measure.
	Probed at the real font and the real 520px column: 28, 30, 32, 34 and
	36 all give three lines; 40 gives four. The name is the longest in the
	catalogue's usual shape -- make-plus-model-plus-fitment -- so a size
	that holds it at three holds most of them.

	`clamp(28px, 2.6vw, 36px)` rather than a flat 36: the rail is capped at
	556 and so keeps its measure from about 1300 up, where 2.6vw has
	already reached the ceiling -- but below that the rail narrows and the
	title has to come down with it or the name runs to four and five lines.
	At 1090 it computes 28, which is where it started.

	THE PRICE IS LEFT ALONE at 32. Dropping it as well would flatten both;
	one of the two had to move and the name is the one that was wrong.
*/
@media (min-width: 1001px) {

	/*
		THE BLOCK CLASS IS IN THE SELECTOR because single-product.css sets
		`font-size: 28px` on `.sf-pdp__rail h1.wp-block-post-title` -- (0,2,1)
		against a bare `.sf-pdp__rail h1` at (0,1,1). Measured with one class:
		the rule was present, matching, and the title still computed 28px.
	*/
	.sf-pdp__rail h1.wp-block-post-title {
		/*
			32, DOWN FROM 36 (operator, the same day: "not nice"). 36 was the
			largest size that still set this name in three lines, and taking
			the ceiling rather than the line count turned out to be the wrong
			end to optimise -- it led the price but dominated the rail. 32
			still leads the 32px price on weight and case alone, and it is
			still three lines.
		*/
		font-size: clamp(28px, 2.6vw, 32px);
		line-height: 1.04;
		/*
			-.02em, up from the -.01em it had. Archivo Black is a heavy face
			and three lines of it at 1.04 want their counters closed a
			little or the block reads as loose beside the price under it.
		*/
		letter-spacing: -.02em;
	}
}


/* ------------------------------------------------------------------ */
/*  The price is set in the display face, not the label face           */
/* ------------------------------------------------------------------ */
/*
	Operator, 2026-11-19: "please make it perfect..".

	**CHAKRA PETCH IS THE LABEL FACE AND THIS WAS THE ONLY PLACE IT WAS
	SET LARGE.** Measured across the rail, every other use of it is a
	small, tracked-out label: the card meta at 10px, FITS at 11, the
	accordion heads and Add to cart at 12, the stock line at 12, the
	quantity at 13. The price sat at 32px in the same face with POSITIVE
	tracking, which is a label habit -- so the most important number on
	the page read as a technical readout rather than as a price.

	INTER 700, BECAUSE THAT IS WHAT THE SAME NUMBER IS SET IN ON THE SHOP.
	Measured on /shop/: the card price is Inter 16px w700. A customer sees
	RM345.00 on the card and then again here; it should be the same face
	in both places. Inter is already loaded on this page -- the related
	band's heading uses it -- so this costs no request.

	NOT Archivo Black, which is the other candidate because it is the
	product NAME's face directly above. At 32px a second block of that
	weight competes with the title rather than following it; the name
	should lead and the price should answer.

	THE COLOUR IS LEFT ALONE. The card's price is muted because a card is
	mostly photograph; here it is a primary fact and takes the full ink.
	Only the face, the weight and the tracking move.

	`.woocommerce-Price-amount` is WooCommerce's own class and the rule is
	scoped to the rail, so the cart, the mini-cart and the archive are
	untouched.
*/
.sf-pdp__rail .wp-block-woocommerce-product-price .woocommerce-Price-amount,
.sf-pdp__rail .wp-block-woocommerce-product-price .woocommerce-Price-amount bdi {
	font-family: 'Inter', var(--sp-body), system-ui, sans-serif;
	font-weight: 700;
	/* Inter at display size wants its counters closed slightly; the
	   +0.5px it had was Chakra Petch's label tracking carried over */
	letter-spacing: -.01em;
}

/* ------------------------------------------------------------------ */
/*  Five empty stars are not a rating                                  */
/* ------------------------------------------------------------------ */
/*
	Operator, 2026-11-19, of the title / rating / price block: "not
	nice" -- then, asked which part: the empty stars, the title's size
	and the spacing, all three.

	**AT ZERO REVIEWS THE RAIL DREW FIVE HOLLOW MAGENTA STARS.** They sit
	directly under the product name, in the brand colour, and what they
	actually communicate is "nobody has bought this" -- which is both
	loud and the opposite of what the link beside them is asking for. The
	link stays; the empty row goes.

	THE TEST IS STRUCTURAL, NOT THE LABEL STRING. `pdp-reviews.js` writes
	`aria-label="Not yet rated"` on the span, which would have been the
	obvious hook and would break silently the day that wording changes.
	Checked on a rated product instead: its stars carry `.sf-rv__on`
	paths and an unrated one carries only `.sf-rv__off`. So the rule is
	"hide the row when not one star is filled", which turns itself off
	the moment a review lands, with nothing to deploy.

	A browser without `:has()` keeps the stars -- today's behaviour, which
	is the right way for this to fail.
*/
.sf-pdp__rail .sf-rv__stars:not(:has(.sf-rv__on)) {
	display: none;
}

/* ------------------------------------------------------------------ */
/*  A button you cannot press is still pink                            */
/* ------------------------------------------------------------------ */
/*
	Operator, 2026-11-19, of the out-of-stock Add to cart: "if cannot buy
	i want color pink that if user see its like disabled .. #ec1e79 maybe
	more light ?"

	**IT WAS ALL BUT INVISIBLE, AND THE OPACITY IS WHY.**
	`single-product.css` greys this state (transparent fill, --sp-line
	border, --sp-muted text) -- and `woocommerce-blocktheme.css` then
	fades every `.button.disabled` to **opacity .5**, so what actually
	painted was a .08-alpha hairline around text at half strength on
	carbon. Measured live before this: background `rgba(0,0,0,0)`, border
	`rgba(242,242,244,.16)`, colour `rgb(154,154,159)`, opacity `.5`.

	So `opacity: 1` is not tidying -- without it every value below is
	halved and the state goes straight back to a ghost. The
	wc-variation-selection-needed rule in single-product.css already had
	to state it for the same reason.

	**A WASH, NOT A PALER PILL.** The literal reading of "more light" is
	a lightened fill -- `color-mix(accent, white)` -- and on a near-black
	page that comes out LOUDER than the live button, so the state that
	cannot be used would be the brightest thing in the buy box. The fill
	is the accent at 18% instead: unmistakably pink, a long way from the
	solid pill, and the contrast with the live button is the fill rather
	than the hue.

	The fill is rgb(56,19,38).

	**THE LABEL IS WHITE** (operator, same day: "text color white"). It was
	`color-mix(accent 70%, white)` -- a light pink, for a monochrome pill --
	and that was the one part of this that was arguing with itself: the
	surface says "off" and the label does not need to say it twice.

	White measures **16.4:1** on this fill, against **4.18:1** for the same
	white on the live pill -- so the state you cannot use now has the more
	legible label of the two. That is the right way round for a line that
	has to be read rather than pressed, and it is also why the FILL has to
	go on carrying the whole signal. **Do not lighten the fill to match the
	label.**

	A literal rather than `--sp-accent-fg`: that token is pitched against
	the magenta, and this background is not magenta.

	**IT IS FLAT AND BORDERLESS** (operator, same day: "should be z-index 1
	and remove border"). It shipped an hour earlier as a translucent wash
	with a 55% accent border, and a translucent fill composites with what
	is behind it -- so the body's carbon weave ran straight through the
	pill and the button read as a hole rather than as a plate.

	**z-index CANNOT FIX THAT AND THE FILL IS WHAT DID.** Stacking order
	decides what paints over what; it does not make a .18-alpha colour
	opaque. The same 18% is mixed with `--sp-bg` instead of `transparent`,
	which is the OPAQUE EQUIVALENT -- the painted colour is unchanged to
	the byte, rgb(56,19,38) either way, and only the weave stops showing.
	`position: relative; z-index: 1` is kept because it was asked for and
	costs nothing, but it is inert for the look; do not read it as the fix.

	**`border-color: transparent`, NOT `border: 0`.** The base rule sets a
	1px border and this button is `box-sizing: border-box`, so zeroing the
	width would hand those 2px back to the padding box and shift the label.
	Keeping the width and clearing the colour leaves the geometry identical
	-- measured 257x52 before and after -- and the background paints under
	it anyway, so no seam opens at the edge.

	**NO `:hover` VARIANT, ON PURPOSE.** `.disabled:hover` would be
	(0,7,3) -- a tie with the `wc-variation-selection-needed` solid pill
	above, which this file loads after -- so hovering the "pick your
	options first" state would wash it out. It is not needed: this rule
	is (0,6,3) like the base `:hover` and the base `.disabled`, and it is
	the last of the three, so it already wins while the pointer is over
	it. Leave it off.

	That selection-needed state keeps its solid pill at (0,7,3), which is
	right: there something the rider picks WILL make the button work.
	This is for the two states where nothing will -- out of stock, and an
	unavailable variation.
*/
body.single-product .sf-pdp .sf-pdp__rail form.cart button.single_add_to_cart_button.disabled,
body.single-product .sf-pdp .sf-pdp__rail form.cart button.single_add_to_cart_button:disabled {
	opacity: 1;
	position: relative;
	z-index: 1;
	background: color-mix(in srgb, var(--sp-accent) 18%, var(--sp-bg));
	border-color: transparent;
	color: #fff;
	cursor: not-allowed;
}

/* ------------------------------------------------------------------ */
/*  The description's list is the description's paragraph              */
/* ------------------------------------------------------------------ */
/*
	Operator, 2026-11-19, of the Description card: "then why the font not
	samee". **The FAMILY and the WEIGHT were already the same** -- Archivo
	400 in both -- so nothing in the type system was wrong. What differed
	was everything else, measured live on the chain cover (1308):

	|            | the lead `<p>` | the `<ul>` |
	|------------|----------------|------------|
	| size       | 14px           | **16.8px** |
	| leading    | 1.32           | **1.55**   |
	| tracking   | .017em         | **none**   |
	| colour     | `--sp-muted`   | **full ink** |

	**NOTHING STYLED THE LIST AT ALL.** Walking every stylesheet for rules
	matching that `<li>` returned an EMPTY set: `single-product.css` reaches
	`.sf-pdp__rail .wp-block-post-content p` and stops there, so the list
	fell through to the theme's own body type -- 1.05rem at 1.55, full ink.
	It is not a competing rule and there was nothing to out-specify; it is a
	gap.

	So this is a near-certain trap for the next block anybody puts in a
	product description. **Style the CONTENT, not the paragraph** -- a
	heading, a table or a blockquote in an imported description would land
	the same way, and the ones here are only right because no product has
	used them yet.

	**THESE FOUR VALUES MIRROR THAT `p` RULE AND MUST MOVE WITH IT.** They
	are written as literals rather than inherited, because the paragraph
	sets them on itself; pointing the list at a token neither rule declares
	would resolve by load order and fail silently, which is the decision
	this file records throughout. **Change both or neither** -- and if the
	bullets should ever lead over the lead line, change THAT rule so the
	pair still moves together, rather than splitting them again.

	The marker takes `color` from its own `<li>`, so the dots follow with
	nothing extra. Spacing is deliberately untouched: the ask was the type,
	and the list's 19.2px approach is the theme's flow rhythm, shared with
	everything else in the card.
*/
body.single-product .sf-pdp .sf-pdp__rail .wp-block-post-content ul,
body.single-product .sf-pdp .sf-pdp__rail .wp-block-post-content ol,
body.single-product .sf-pdp .sf-pdp__rail .wp-block-post-content li {
	font-size: 14px;
	line-height: 1.32;
	letter-spacing: .017em;
	color: var(--sp-muted);
}

/* ------------------------------------------------------------------ */
/*  The meta values sit with the card, not above it                    */
/* ------------------------------------------------------------------ */
/*
	Operator, 2026-11-19, of Product details and Specifications: "why have
	text white , i think its not nice .. do you have others color that blend
	with that".

	Measured in those two cards, everything 13px Archivo 400:

	| | |
	|---|---|
	| Category: / Tags: / Brands: / Carbon | `--sp-muted` #9a9a9f |
	| Bodykits, Y15ZR, Superfly, Forged Carbon V1 | **`--sp-text` #f2f2f4** |

	So the values were the BRIGHTEST text in the rail -- brighter than the
	description, brighter than the product's own short copy -- and what they
	are is a list of taxonomy terms. They now take **.72 of the same ink**,
	which lands about #b3b3b5 over the page: clearly a step up from the
	labels beside them, no longer the loudest thing in the card.

	**NOT ALL THE WAY DOWN TO `--sp-muted`.** "Category:" and "Bodykits" are
	a label and its value and have to read as different things; one colour
	for both turns each row into a single grey run. The step is what carries
	that, so if this is ever re-pitched, move the VALUE and leave the label
	where it is.

	**THESE ARE LINKS AND THEIR AFFORDANCE IS THE HOVER.** Nothing underlines
	them and nothing colours them, so brightness was the only thing marking
	them -- but `single-product.css` already turns them `--sp-accent` on
	hover, which is the real signal and is untouched. Measured after: 9.2:1
	on the page, against 6.8:1 for the labels, so this is comfortably legible
	and not a contrast trade.

	Both selectors have to clear (0,3,2): the terms are
	`.sf-pdp__rail .wp-block-post-terms a` and the attribute cell is
	`.sf-pdp__card-body .sf-pdp__attrs table.shop_attributes td`, both in
	`single-product.css`.
*/
body.single-product .sf-pdp .sf-pdp__rail .wp-block-post-terms a,
body.single-product .sf-pdp .sf-pdp__rail .sf-pdp__attrs table.shop_attributes td,
body.single-product .sf-pdp .sf-pdp__rail .sf-pdp__attrs table.shop_attributes td p {
	color: rgba(242, 242, 244, .72);
}

/* ------------------------------------------------------------------ */
/*  The delivery and warranty lines are one colour                     */
/* ------------------------------------------------------------------ */
/*
	Operator, 2026-11-19, of the two assurance cards: "why have color white
	.. the text .. why not same color".

	Measured, both rows 13px Archivo:

	| | colour | weight |
	|---|---|---|
	| the sentence | `--sp-muted` #9a9a9f | 400 |
	| "2-4 working days", "6-month warranty" | **`--sp-text` #f2f2f4** | **700** |
	| the /warranty link | **`--sp-text`** | 400 |

	**THE BOLD WAS BEING SAID TWICE.** Those phrases are already 700 against
	a 400 sentence, which is the emphasis; the colour was a second signal
	doing the same job, and on a 13px line inside a small card it read as two
	different texts rather than one. Everything in the row inherits the row's
	colour now and **the weight carries it alone** -- which it still does
	plainly at 700 against 400.

	`color: inherit` rather than naming the token, so if the row is ever
	re-pitched these follow it instead of drifting apart again.

	**THEY ARE `<b>`, NOT `<strong>`.** A probe written against `strong`
	finds nothing here and reports "no rule sets this", which is how an
	earlier pass concluded the colour came from inheritance when it does
	not. Both are named below so the markup can change without this quietly
	stopping.

	**THE LINK KEEPS ITS HOVER.** Dropping /warranty to the row colour takes
	away the only thing marking it as a link apart from its underline, so the
	accent hover is restated here -- `color: inherit` at this specificity
	would otherwise beat the existing hover rule and the link would never
	change at all.
*/
body.single-product .sf-pdp .sf-pdp__rail .sf-pdp__assure-row b,
body.single-product .sf-pdp .sf-pdp__rail .sf-pdp__assure-row strong,
body.single-product .sf-pdp .sf-pdp__rail .sf-pdp__assure-row a {
	color: inherit;
}

body.single-product .sf-pdp .sf-pdp__rail .sf-pdp__assure-row a:hover,
body.single-product .sf-pdp .sf-pdp__rail .sf-pdp__assure-row a:focus-visible {
	color: var(--sp-accent);
}
