@nurix/etna 0.19.66 → 0.19.68

package/data/harnesses/apollo/skills/design/data/elevation/material-3-tonal-elevation.md

@@ -1,21 +0,0 @@
----
-id: material-3-tonal-elevation
-class: a3
-sourced: 2026-07-12
-source: "distilled from m3.material.io/styles/elevation/tokens; shadow figures approximate Material Web's published elevation-level tokens"
-license: open
-feel: "quiet tech-forward calm; depth reads as a warm color shift more than a cast shadow, corners softly rounded"
----
-Material 3's headline move: elevation is expressed primarily through **tonal surface tint** (a translucent overlay of the primary brand color, stronger at higher levels) with a thin shadow underneath as a secondary cue. No borders — surfaces separate by tone, not line. Best where a single brand hue should visibly "light up" whatever floats highest.
-
-Approximation note: shadow figures approximate Material Web's published elevation-level tokens — check exact rgba/opacity figures against the live source before shipping.
-
-```yaml values
-radius: { sm: 8, md: 12, lg: 16 }
-hairline: none
-shadow:
-  level1: "0px 1px 2px rgba(0,0,0,0.30), 0px 1px 3px 1px rgba(0,0,0,0.15) + {primary-hue tint / ~5% overlay}"
-  level2: "0px 1px 2px rgba(0,0,0,0.30), 0px 2px 6px 2px rgba(0,0,0,0.15) + {primary-hue tint / ~8% overlay}"
-  level3: "0px 4px 8px 3px rgba(0,0,0,0.15), 0px 1px 3px rgba(0,0,0,0.30) + {primary-hue tint / ~11% overlay}"
-dark-ladder: { step: "+~3% primary-tint per level (approx 5/8/11/12/14%)", hue: drift }
-```

package/data/harnesses/apollo/skills/design/data/elevation/neobrutalist-hard-offset.md

@@ -1,19 +0,0 @@
----
-id: neobrutalist-hard-offset
-class: a3
-sourced: 2026-07-12
-source: "distilled from neubrutalism.com and community neobrutalism CSS references (community pattern, no single owning org)"
-license: open
-feel: "loud, tactile, intentionally raw; a solid offset shadow with zero blur reads as a physical card lifted straight off the page"
----
-Neobrutalism inverts the entire elevation vocabulary: no blur, ever. A thick (3–4px) solid black border plus a flat, hard-edged offset shadow in pure black (or a bold flat color) does all the depth work — the shadow is a solid silhouette displaced down-right, not a soft cast. Radius stays at or near zero to keep the "raw" read.
-
-```yaml values
-radius: { sm: 0, md: 0, lg: 4 }
-hairline: "not a hairline — a 3-4px solid black border replaces the border/shadow distinction entirely"
-shadow:
-  sm: "4px 4px 0px #000000"
-  md: "6px 6px 0px #000000"
-  lg: "8px 8px 0px #000000"
-dark-ladder: { step: "n/a — most kits binary-invert black/white rather than laddering surfaces", hue: held }
-```

package/data/harnesses/apollo/skills/design/data/elevation/neumorphic-soft-dual-shadow.md

@@ -1,18 +0,0 @@
----
-id: neumorphic-soft-dual-shadow
-class: a3
-sourced: 2026-07-12
-source: "distilled from neumorphism.io and community soft-UI CSS references (community pattern, no single owning org)"
-license: open
-feel: "soft, extruded, monochrome; elements look pressed into or popped out of a background that shares their exact color"
----
-Neumorphism fakes physical extrusion with a **paired light+dark shadow** on an element whose background exactly matches its parent's — no border, no color contrast, only two offset shadows (one lighter, one darker than the shared base) doing all the modeling. Pressed/inset states simply invert the pair. Contrast-fragile by construction: raising shadow strength enough to pass accessibility checks destroys the soft illusion that is the entire point.
-
-```yaml values
-radius: { sm: 12, md: 16, lg: 24 }
-hairline: "none — a drawn border would break the illusion; edges exist only via the shadow pair"
-shadow:
-  raised: "10px 10px 20px #bebebe, -10px -10px 20px #ffffff"
-  pressed: "inset 6px 6px 12px #bebebe, inset -6px -6px 12px #ffffff"
-dark-ladder: { step: "n/a — dark variant is a manual per-surface retune (e.g. base #2d3239, shadows #23272d / #373d46), not a formula", hue: held }
-```

package/data/harnesses/apollo/skills/design/data/elevation/notion-warm-minimal-hairline.md

@@ -1,17 +0,0 @@
----
-id: notion-warm-minimal-hairline
-class: a3
-sourced: 2026-07-12
-source: "distilled from github.com/VoltAgent/awesome-design-md — notion/DESIGN.md"
-license: reference-only
-feel: "warm, paper-like minimalism; content over chrome, separation is almost always a thin line rather than a cast shadow"
----
-Notion's house rule is explicit: don't apply heavy shadows on flat documentation cards. Separation is carried almost entirely by a single hairline border token reused across nav, cards, and containers; the rare exception is a very soft lift on preview/mockup cards, kept subtle enough not to compete with content.
-
-```yaml values
-radius: { sm: 4, md: 6, lg: 10 }
-hairline: "1px solid {colors.hairline} #EDEEEC — the default separator for nav, cards, and containers; heavy shadows explicitly avoided"
-shadow:
-  rare-lift: "used sparingly, only on mockup/preview cards — e.g. 0 4px 16px rgba(0,0,0,.06)"
-dark-ladder: { step: "+~2-3L per level (approx #191919 → #202020 → #2f2f2f surfaces)", hue: held }
-```

package/data/harnesses/apollo/skills/design/data/elevation/polaris-elevation-inset-bevel.md

@@ -1,21 +0,0 @@
----
-id: polaris-elevation-inset-bevel
-class: a3
-sourced: 2026-07-12
-source: "distilled from polaris-react.shopify.com/design/depth/shadow-tokens; radius px approximated, exact figures not published in source"
-license: open
-feel: "commerce-admin friendly; raised surfaces read as physically thick, machined edges rather than flat cards floated on a shadow"
----
-Polaris splits shadow into three distinct token families used in combination: elevation (the classic cast drop-shadow), inset (an embedded/pressed look), and bevel (a thin inner highlight+shadow edge that gives a raised surface the sense of physical thickness). Combining bevel with elevation is the house move for "this is a real, liftable object."
-
-Approximation note: radius px approximated — exact figures are not published in the source and should be checked before shipping.
-
-```yaml values
-radius: { sm: 4, md: 8, lg: 12 }
-hairline: "implicit — bevel-100 supplies a 1px inner highlight/shadow edge in place of a plain border"
-shadow:
-  elevation-100: "cast shadow, lightest — resting cards"
-  elevation-300: "cast shadow, medium — popovers and overlays"
-  bevel-100: "inset highlight + inset shadow — thick machined edge on raised surfaces"
-dark-ladder: { step: "n/a — Shopify admin ships no official dark theme", hue: held }
-```

package/data/harnesses/apollo/skills/design/data/elevation/radix-themes-alpha-ring-scale.md

@@ -1,19 +0,0 @@
----
-id: radix-themes-alpha-ring-scale
-class: a3
-sourced: 2026-07-12
-source: "distilled from radix-ui.com/themes/docs/theme/shadows, /radius, and the shadow.css token source"
-license: open
-feel: "quiet, correct, slightly clinical; a 6-step alpha ramp any accent color can flow through without redoing the shadows"
----
-Radix builds its shadow ramp entirely from **gray-alpha and black-alpha overlays** rather than fixed opaque colors, so the same six shadow steps stay correct across every accent and every light/dark theme automatically. `shadow-1` is an inset ring that doubles as the hairline for flush, level-0 surfaces.
-
-```yaml values
-radius: { sm: 4, md: 8, lg: 12 }
-hairline: "shadow-1 — inset 0 0 0 1px var(--gray-a5), inset 0 1.5px 2px var(--gray-a2)/var(--black-a2)"
-shadow:
-  shadow-2: "0 0 0 1px var(--gray-a3), 0 0 0 .5px var(--black-a1), 0 1px 1px var(--gray-a2), 0 2px 1px -1px var(--black-a1), 0 1px 3px var(--black-a1)"
-  shadow-3: "0 0 0 1px var(--gray-a3), 0 2px 3px -2px var(--gray-a3), 0 3px 12px -4px var(--black-a2), 0 4px 16px -8px var(--black-a2)"
-  shadow-5: "0 0 0 1px var(--gray-a3), 0 12px 60px var(--black-a3), 0 12px 32px -16px var(--gray-a5)"
-dark-ladder: { step: "n/a — alpha-step ramp (gray-a3 → a5 → a7 in dark), not a lightness ladder", hue: held }
-```

package/data/harnesses/apollo/skills/design/data/elevation/shadcn-neutral-derived-radius.md

@@ -1,19 +0,0 @@
----
-id: shadcn-neutral-derived-radius
-class: a3
-sourced: 2026-07-12
-source: "distilled from ui.shadcn.com/docs/theming; shadow values are the Tailwind defaults shadcn themes typically inherit"
-license: open
-feel: "neutral, restrained utility feel; nearly flat, with one hairline border doing most of the separating work"
----
-shadcn/ui derives its entire radius scale from a single `--radius` variable (multiplied up/down for sm/md/lg/xl), and leans on a plain neutral `border` token rather than shadow for everyday separation — shadow is reserved for popovers/menus. The whole system is intentionally colorless so any brand palette can sit on top of it unchanged.
-
-```yaml values
-radius: { sm: 6, md: 8, lg: 10 }
-hairline: "border token — 1px solid, neutral gray/oklch; default separator for card, input, and popover"
-shadow:
-  sm: "0 1px 2px 0 rgb(0 0 0 / 0.05)"
-  md: "0 4px 6px -1px rgb(0 0 0 / 0.1), 0 2px 4px -2px rgb(0 0 0 / 0.1)"
-  popover: "0 10px 15px -3px rgb(0 0 0 / 0.1), 0 4px 6px -4px rgb(0 0 0 / 0.1)"
-dark-ladder: { step: "+0.03–0.05 L per surface (oklch background → card)", hue: held }
-```

package/data/harnesses/apollo/skills/design/data/elevation/stripe-brand-tinted-dual-shadow.md

@@ -1,18 +0,0 @@
----
-id: stripe-brand-tinted-dual-shadow
-class: a3
-sourced: 2026-07-12
-source: "distilled from designmd.run/blog/stripe-design-system-breakdown and docs.stripe.com/elements/appearance-api"
-license: reference-only
-feel: "warm, trustworthy, faintly atmospheric; shadows carry Stripe's navy-purple brand hue instead of reading as neutral gray"
----
-Stripe's shadows are two stacked layers where the far/ambient layer is tinted `rgba(50,50,93,...)` — the same blue-gray family as its navy-purple brand palette — with only the near/contact layer left as neutral black. Radius is deliberately conservative (Elements' own default is 4px) — Stripe treats restraint on both shadow and radius as part of feeling "trustworthy."
-
-```yaml values
-radius: { sm: 4, md: 6, lg: 8 }
-hairline: "none specified — separation relies on shadow against a cool-gray canvas (#F6F9FC)"
-shadow:
-  near: "rgba(0,0,0,.1) 0px 1px 2px 0px"
-  card: "rgba(50,50,93,.25) 0px 6px 12px -2px, rgba(0,0,0,.3) 0px 3px 7px -3px — {navy-purple tint / 0.25 alpha on far layer}"
-dark-ladder: { step: "unconfirmed — Stripe's public brand system is light-first with no published dark ramp", hue: held }
-```

package/data/harnesses/apollo/skills/design/data/elevation/tailwind-default-utility-scale.md

@@ -1,19 +0,0 @@
----
-id: tailwind-default-utility-scale
-class: a3
-sourced: 2026-07-12
-source: "distilled from tailwindcss.com/docs/box-shadow default theme values"
-license: open
-feel: "the internet's default; a neutral gray shadow ramp so common it reads as 'no design system' unless overridden"
----
-Tailwind's out-of-the-box shadow scale is the de facto baseline for a huge share of the web: six pure-neutral steps (sm → 2xl) with no brand tint, used independently of its separate border-color utilities rather than paired with them. Its very ubiquity is the point of including it here — it's the "control group" every more opinionated system in this list is reacting against.
-
-```yaml values
-radius: { sm: 4, md: 6, lg: 8 }
-hairline: "border-gray-200 (or theme border color), 1px solid — used independently of the shadow scale, not paired with it"
-shadow:
-  sm: "0 1px 2px 0 rgb(0 0 0 / 0.05)"
-  md: "0 4px 6px -1px rgb(0 0 0 / 0.1), 0 2px 4px -2px rgb(0 0 0 / 0.1)"
-  xl: "0 20px 25px -5px rgb(0 0 0 / 0.1), 0 8px 10px -6px rgb(0 0 0 / 0.1)"
-dark-ladder: { step: "n/a — no shipped dark ramp; convention is dark:shadow-none or a manual opacity boost", hue: held }
-```

package/data/harnesses/apollo/skills/design/data/elevation/vercel-geist-shadow-as-border.md

@@ -1,20 +0,0 @@
----
-id: vercel-geist-shadow-as-border
-class: a3
-sourced: 2026-07-12
-source: "distilled from vercel.com/geist/materials"
-license: reference-only
-feel: "razor-flat product chrome; the resting 'border' is secretly a zero-blur shadow, softness saved entirely for menus and modals"
----
-Geist's signature trick: a zero-offset, zero-blur, 1px-spread shadow (`0 0 0 1px rgba(0,0,0,.08)`) stands in for a border at rest, avoiding box-model side effects. As elements go from card → menu → modal, additional shadow layers stack on top of that same base line rather than replacing it, so the "border" never disappears.
-
-```yaml values
-radius: { sm: 6, md: 12, lg: 16 }
-hairline: "shadow-emulated hairline — 0px 0px 0px 1px rgba(0,0,0,.08), zero blur, present at every rest state"
-shadow:
-  base: "0px 0px 0px 1px rgba(0,0,0,.08)"
-  small: "0px 0px 0px 1px rgba(0,0,0,.08), 0px 2px 4px rgba(0,0,0,.04)"
-  menu: "0px 0px 0px 1px rgba(0,0,0,.08), 0px 8px 24px -8px rgba(0,0,0,.12)"
-  modal: "0px 0px 0px 1px rgba(0,0,0,.08), 0px 16px 48px -12px rgba(0,0,0,.18)"
-dark-ladder: { step: "+1 surface swap per level (#000 → #0a0a0a → #111), occasional {blue-hue tint / low-alpha} wash surface", hue: drift }
-```

package/data/harnesses/apollo/skills/design/data/iconography.md

@@ -1,199 +0,0 @@
----
-id: iconography
-class: a5
-sourced: 2026-07-12
-source: "distilled from the A5 iconography phase-1 research sweep — Lucide, Tabler Icons, Radix Icons, Heroicons, Phosphor (Regular + Duotone), Material Symbols, IBM Carbon, Octicons, Linear, and Stripe Apps"
-license: mixed — nine of the eleven conventions below are open (OSS icon sets under ISC/MIT/Apache-2.0); two are licensed (Linear, Stripe) because they describe a proprietary product's house convention rather than a redistributable asset — see the `license` line inside each convention's own values block
-feel: "the register this file spans: quiet neutral utility (Lucide, Tabler) through dense-precision (Radix), paired-weight (Heroicons), rounded-friendly and illustrative (Phosphor), mechanically variable (Material Symbols), rectilinear enterprise (Carbon), hand-tuned developer precision (Octicons), to near-monochrome restraint (Linear, Stripe) — pick the register the product needs, not just a glyph shape"
----
-A small conventions database, not a big catalog — eleven curated iconography conventions, each a coherent bundle of icon set, stroke/fill treatment, size grid, default color role, and type-pairing rule, because real products never choose these independently. The standing complaint every entry below answers is the same one: a bare, icon-less nav reads as unfinished; each convention states plainly where icons are obligatory (nav, status glyphs, empty states) versus optional (body copy, pure decoration).
-
-## Convention: lucide-quiet-utility
-
-**Feel:** "quiet, neutral utility line-work — reads as structural chrome, not decoration"
-
-Lucide (ISC-licensed fork of Feather) is the default pick for stroke-based quiet utility icons in SaaS dashboards. Obligatory in nav rails, sidebar items, and empty-state graphics — the failure mode this convention guards against is exactly a bare, icon-less nav; optional inline in body copy or button labels where the text alone already carries the meaning.
-
-```yaml values
-set: Lucide (ISC license)
-stroke: "2px, currentColor, round caps/joins, no fills"
-grid: [24, 24]
-default-color: muted-foreground
-pairing: "absoluteStrokeWidth keeps the 2px line constant across sizes; optically matches a 14-16px UI sans at regular weight"
-obligatory: [nav-rail, sidebar-items, empty-state-graphics]
-optional: [body-copy, button-labels]
-license: open
-```
-
-## Convention: tabler-technical-outline
-
-**Feel:** "sharp, technical, exhaustive — the 'there's an icon for everything' catalog reads as engineering completeness, not a curated voice"
-
-Tabler's 6,000+ MIT icon catalog is the pick when coverage matters more than a distinct personality — admin panels, internal tools, anywhere a designer needs an icon for a niche concept without commissioning custom art. Obligatory in dense toolbars and table row-actions where recognition beats charm; optional as pure decoration, where its neutrality reads generic rather than intentional.
-
-```yaml values
-set: Tabler Icons (MIT license)
-stroke: "2px, adjustable, currentColor"
-grid: [24, 24]
-default-color: "neutral-500 / muted-foreground"
-pairing: "2px stroke reads correctly against 14px UI text; bump to ~2.5px only above 32px display sizes so the line doesn't thin out"
-obligatory: [dense-toolbars, table-row-actions]
-optional: [pure-decoration]
-license: open
-```
-
-## Convention: radix-dense-precision
-
-**Feel:** "crisp, hard-edged precision built for the smallest interactive targets — dropdowns, toolbars, table cells, badges"
-
-Radix Icons' 15x15 grid (not the usual 24) exists because at the sizes dense UI actually renders icons — inside buttons, selects, table cells — a 24-grid icon scaled down loses pixel-snap crispness, while a native 15px drawing keeps strokes on-pixel. Obligatory wherever the paired control is itself sub-20px (combobox triggers, sort carets, badge glyphs); skip above ~20px, where the fixed grid stops paying off and a 24-grid set reads better — there is no "optional" tier here, only a hard size threshold.
-
-```yaml values
-set: "Radix Icons (MIT license, designed by WorkOS)"
-stroke: "single fixed weight per icon — redrawn per size, not algorithmically scaled"
-grid: [15, 15]
-default-color: "gray-11 / muted-foreground"
-pairing: "matches the cap-height of 12-13px UI text used across Radix-based component libraries (Radix Themes, shadcn primitives)"
-obligatory: [combobox-triggers, sort-carets, badge-glyphs]
-optional: []
-skip-above: "~20px — hand off to a 24-grid set past this threshold"
-license: open
-```
-
-## Convention: heroicons-outline-solid-pair
-
-**Feel:** "Tailwind-native default — outline for resting state, solid for the moment something becomes true"
-
-Heroicons ships two coherent weights meant to be switched, never mixed within one context: 24px/1.5px-stroke outline for default and primary nav, 20px solid (plus a 16px mini and micro variant) for active/selected states and compact controls. Obligatory swap pattern: an outline icon in a nav item becomes its solid twin exactly when that item is current — the fill change is the only state signal, so don't blend outline and solid within the same list.
-
-```yaml values
-set: "Heroicons (MIT license, Tailwind Labs)"
-stroke: "1.5px outline / filled solid — paired variants, not a weight slider"
-grid: [24, 24] # outline; solid ships at 20px, mini at 16px, micro at 16px
-default-color: "text-gray-500 (outline, resting) -> text-gray-900 or accent (solid, active)"
-pairing: "built on Tailwind's own h-4/h-5/h-6 scale, so icon size always lands exactly on a text-size step"
-obligatory: [nav-active-state-swap]
-optional: []
-license: open
-```
-
-## Convention: phosphor-rounded-friendly
-
-**Feel:** "rounded, approachable, slightly playful without tipping into cartoonish — the friendly-SaaS register between Lucide's neutrality and duotone's illustration"
-
-Phosphor's Regular weight is the rounded-friendly middle ground: softer terminals than Lucide or Tabler, but still a line icon, not a shape. Obligatory in consumer-facing or community products (onboarding flows, empty states, nav) where the brand voice is warm rather than clinical; match icon weight to type weight — Regular icon beside regular/medium body text, Bold icon only beside semibold-or-heavier headings, so the icon never visually undershoots its label.
-
-```yaml values
-set: "Phosphor Icons — Regular weight (MIT license)"
-stroke: "~1.5px-equivalent stroke, rounded caps and joins"
-grid: [24, 24] # authored on a 256x256 source grid, shipped at 24px
-default-color: "muted-foreground, warms to accent on hover/active"
-pairing: "weight axis tracks type weight: Regular <-> body/UI text, Bold <-> semibold+ headings"
-obligatory: [onboarding-flows, empty-states, nav]
-optional: []
-license: open
-```
-
-## Convention: phosphor-duotone-expressive
-
-**Feel:** "filled, two-tone, illustrative — reads as a fragment of an illustration more than a clickable control"
-
-Phosphor's Duotone weight layers a full-opacity foreground shape over a 20%-opacity fill of the same color, giving flat icons a sense of depth without a second hue. Obligatory in empty states, onboarding checklists, and marketing-adjacent surfaces where an icon needs to carry emotional weight; never use duotone for functional nav or toolbar icons — the two-tone fill reads as illustration rather than an actionable control, and it undermines click affordance.
-
-```yaml values
-set: "Phosphor Icons — Duotone weight (MIT license)"
-stroke: "filled: 100% opacity primary shape + 20% opacity secondary layer, same hue"
-grid: [24, 24]
-default-color: "brand/accent (duotone is inherently expressive, rarely neutral gray)"
-pairing: "sized 1.5-2x the adjacent display headline's cap height; never paired directly with dense UI text"
-obligatory: [empty-states, onboarding-checklists, marketing-adjacent-surfaces]
-optional: []
-never-use: [functional-nav, toolbar-icons]
-license: open
-```
-
-## Convention: material-symbols-variable-axes
-
-**Feel:** "mechanically consistent, endlessly adjustable — one variable font file standing in for what other systems ship as five separate icon sets"
-
-Material Symbols collapses weight, fill, grade, and optical size into a single variable font (`wght` 100-700, `FILL` 0/1, `GRAD` -50 to 200, `opsz` 20-48), so "outline vs. filled" and "thin vs. bold" become live CSS values instead of separate asset swaps. Obligatory pattern: `FILL 0` for resting nav items, animated to `FILL 1` on select — the fill toggle is the selected-state affordance itself, so a nav that skips it loses its only state signal.
-
-```yaml values
-set: "Material Symbols (Apache-2.0 license, Google)"
-stroke: "variable FILL axis: 0 (outlined) to 1 (filled); wght 100-700, GRAD -50 to 200"
-grid: [24, 24] # opsz axis spans 20-48 on the same glyph
-default-color: "on-surface-variant (Material's muted-foreground equivalent)"
-pairing: "opsz should numerically track the paired text's rendered px/pt size 1:1, or stroke weight looks wrong at the range's edges"
-obligatory: [nav-fill-toggle-as-selected-state]
-optional: []
-license: open
-```
-
-## Convention: carbon-enterprise-grid
-
-**Feel:** "rectilinear, high-density, unmistakably enterprise — icons read as instrumentation, not ornament"
-
-IBM Carbon draws every icon on a 32x32 master grid with 2px padding, then scales linearly down to 24/20/16px rather than redrawing per size, keeping optical alignment exact across a data-dense UI. Obligatory at 16px inside buttons and inputs, where IBM Plex Sans runs at 14px; 24/32px reserved for standalone actions and empty-state illustration slots. Touch targets stay 44px+ regardless of the visual icon size — a Carbon-specific accessibility floor worth keeping alongside the visual convention.
-
-```yaml values
-set: "IBM Carbon icons (Apache-2.0 license)"
-stroke: "solid/filled glyph forms scaled from one master grid, not a stroke-weight system"
-grid: [32, 32] # scales down linearly to 24/20/16 for actual component use
-default-color: "icon-primary token (near-black in light theme, near-white in dark)"
-pairing: "16/20px icons pair with 14/16px IBM Plex Sans UI text; 24/32px reserved for standalone or empty-state use"
-obligatory: [buttons-and-inputs-at-16px, standalone-actions-at-24-32px, empty-state-illustration]
-optional: []
-touch-target-floor: "44px+ regardless of visual icon size"
-license: open
-```
-
-## Convention: octicons-dual-precision
-
-**Feel:** "hand-tuned developer-tool precision — every icon individually redrawn per size rather than mechanically scaled"
-
-GitHub's Octicons draws each icon twice — a 16px version and a separately hand-corrected 24px version, rather than one scaled from the other — with a consistent 1.5px stroke, round caps/joins, and 1px corner radius across both. Obligatory at 16px in repo/PR/issue navigation chrome, where an icon-less nav row reads as visibly broken against GitHub's own conventions; 24px reserved for section headers and empty-state glyphs, never mixed with 16px in the same row.
-
-```yaml values
-set: "Octicons — GitHub Primer (MIT license)"
-stroke: "1.5px, round caps and joins, 1px corner radius (0.25px on small filled arrowheads)"
-grid: [16, 16] # plus a separately hand-drawn 24x24 variant
-default-color: "fg.muted (Primer token)"
-pairing: "16px icon pairs with Primer's 14px UI text baseline; 24px pairs with section headers only"
-obligatory: [repo-pr-issue-nav-chrome-at-16px, section-headers-and-empty-states-at-24px]
-optional: []
-license: open
-```
-
-## Convention: linear-icon-restraint
-
-**Feel:** "near-monochrome restraint — icons are load-bearing wayfinding, never decoration, and the restraint itself is the brand voice"
-
-Linear confines icons almost exclusively to two obligatory slots — the sidebar nav rail and issue/status glyphs (priority, workflow state) — and deliberately withholds them everywhere else (body text, secondary inline actions) to protect density. The one place Linear over-indexes on icons rather than under-indexing is exactly where the standing complaint bites hardest: a nav with missing icons reads as unfinished, so nav iconography is treated as non-negotiable even though icons everywhere else stay optional.
-
-```yaml values
-set: "product convention (implementation-agnostic thin outline, Lucide-class)"
-stroke: "~1.5px, single weight, no fill states except the active accent swap"
-grid: [16, 16]
-default-color: "muted-foreground, promotes to the single accent color only on active/selected state"
-pairing: "icon optical size matches the 13px UI sans cap height in the sidebar; icon and label baseline-align, never icon-only in primary nav"
-obligatory: [sidebar-nav-rail, issue-status-glyphs]
-optional: [body-text, secondary-inline-actions]
-license: licensed
-```
-
-## Convention: stripe-restrained-fintech
-
-**Feel:** "near-monochromatic, trustworthy, financial-instrument register — icons are quiet infrastructure, color is spent on one accent only"
-
-Stripe's Icon component ships a fixed five-step size scale (xsmall through xlarge) rather than free pixel values, keeping every icon on-grid with the dashboard's type scale by construction; icons default to inherit surrounding text color and only break to the brand indigo for strokes that need to signal interactivity. Obligatory in the dashboard nav rail and the app-indicator badge (the colored bar + icon identifying which Stripe App is active); optional-to-discouraged as decoration inside data-dense tables, where Stripe's own guidance limits custom styling to preserve platform consistency.
-
-```yaml values
-set: "product convention (Stripe Dashboard / Stripe Apps Icon component)"
-stroke: "thin outline, inherits currentColor by default"
-grid: [16, 20, 24, 28, 32] # xsmall..xlarge step scale, not free px
-default-color: "inherits surrounding text color (currentColor); accent indigo for interactive strokes only"
-pairing: "icon size step always matches the adjacent text's type-scale step, never set independently"
-obligatory: [dashboard-nav-rail, app-indicator-badge]
-optional: [decoration-in-dense-tables]
-license: licensed
-```

package/data/harnesses/apollo/skills/design/data/motion/accordion-expand.md

@@ -1,18 +0,0 @@
----
-id: accordion-expand
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/animation-vocabulary/SKILL.md (Accordion/Collapse term) + review-animations/STANDARDS.md (easing decision order — on-screen morph → strong ease-in-out; interruptibility)"
-license: open
-feel: "a 240ms reveal on a strong ease-in-out reads smooth; snapping open or drifting past 300ms reads clunky"
-aliases: { clunky: "a plain built-in easing instead of the strong custom ease-in-out curve", "snap-open": "no transition at all — content appears/disappears instantly", janky: "keyframes (or animated height) instead of an interruptible transition on transform/clip-path" }
----
-Expand/collapse sections, FAQ rows, nested tree rows — an on-screen morph, so ease-in-out, not the entrance curve. Prefer `clip-path`/transform reveals over animated `height` (layout property — off-GPU). Rapid toggling must retarget mid-animation, not restart from collapsed.
-
-```yaml values
-duration: 240ms
-easing: cubic-bezier(0.77, 0, 0.175, 1)
-element-class: panel
-interruptible: required
-budget: { panel: 200-300ms }
-```

package/data/harnesses/apollo/skills/design/data/motion/destructive-hold-confirm.md

@@ -1,18 +0,0 @@
----
-id: destructive-hold-confirm
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/review-animations/STANDARDS.md (asymmetric timing — press slow/release fast example, values verbatim class)"
-license: open
-feel: "a slow deliberate fill on hold with a fast release reads like the system is making you mean it; a symmetric hold/release reads accidental-prone"
-aliases: { "trigger-happy": "hold and release animating at the same speed, so a brief accidental hold nearly commits", "accidental-prone": "no hold gate at all — a single tap commits the destructive action" }
----
-Delete/destructive confirmations gated behind a hold, not a second click. The hold phase (user deciding) animates slow and must be abortable mid-hold; release or completion (system responding) snaps back fast.
-
-```yaml values
-duration: 2000ms (hold-to-fill) / 200ms (release)
-easing: linear (hold) / cubic-bezier(0.23, 1, 0.32, 1) (release)
-element-class: control
-interruptible: required
-budget: { control: 1500-2500ms hold, 150-250ms release }
-```

package/data/harnesses/apollo/skills/design/data/motion/drag-dismiss-momentum.md

@@ -1,18 +0,0 @@
----
-id: drag-dismiss-momentum
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/review-animations/STANDARDS.md (gestures & drag — velocity dismissal; springs config) + apple-design/SKILL.md §5 (velocity handoff)"
-license: open
-feel: "a flick that keeps its exact velocity into the dismiss reads thrown away on purpose; a fixed-duration tween to the edge reads like the drag didn't matter"
-aliases: { disconnected: "velocity discarded at release in favor of a fixed-duration keyframe to the exit point", stuck: "dismissal gated on a distance threshold instead of release velocity (~0.11 px/ms)" }
----
-Swipe-to-dismiss on toasts, cards, sheets. The spring here is a correctness mechanism (velocity continuity for gesture-driven motion), not styling — it carries the pointer's release velocity into the exit and stays grabbable and reversible mid-flight. Bounce is canon-sanctioned specifically for drag-to-dismiss; it doesn't generalize.
-
-```yaml values
-duration: 500ms (spring settle; no fixed duration)
-easing: spring(bounce 0.2, damping ~0.8)
-element-class: overlay
-interruptible: required
-budget: { overlay: 350-550ms }
-```

package/data/harnesses/apollo/skills/design/data/motion/empty-state-arrive.md

@@ -1,18 +0,0 @@
----
-id: empty-state-arrive
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/review-animations/STANDARDS.md (springs — subtle bounce) + improve-animations/AUDIT.md §1 (rare moments); A6 admission law (identity-moment scope)"
-license: open
-feel: "an illustration that gently scales and settles in reads intentional, like the empty state was designed on purpose; a plain instant fade reads unfinished"
-aliases: { unfinished: "the empty-state illustration/copy just appearing with no entrance at all", overdone: "spring bounce strong enough to feel jokey rather than reassuring" }
----
-First-visit or true-zero-data empty states (no rows yet, no results yet). Low-frequency identity moment — springy motion is admissible, but keep bounce subtle since the tone is reassuring, not playful.
-
-```yaml values
-duration: 450ms (spring; no fixed duration)
-easing: spring(bounce 0.15)
-element-class: identity-moment
-interruptible: n/a
-budget: { identity-moment: 350-550ms }
-```

package/data/harnesses/apollo/skills/design/data/motion/hover-lift.md

@@ -1,18 +0,0 @@
----
-id: hover-lift
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/emil-design-eng/SKILL.md (easing decision — hover → ease) + review-animations/STANDARDS.md (frequency tiers; hover gating); Easing Blueprint agrees as secondary"
-license: open
-feel: "a soft `ease` color/elevation shift under 200ms reads responsive; past 250ms or on touch it reads laggy and fake"
-aliases: { laggy: "hover duration pushed past the control budget", "fake-hover": "hover motion firing on a touch tap because it isn't gated to hover:hover and pointer:fine" }
----
-List rows, nav items, cards, toolbar icons on mouse-hover. Gate behind `@media (hover: hover) and (pointer: fine)` always. Sits in the "tens of times/day — reduce, don't eliminate" frequency tier; drop it first on any list scrolled constantly.
-
-```yaml values
-duration: 150ms
-easing: ease
-element-class: control
-interruptible: n/a
-budget: { control: 120-200ms }
-```

package/data/harnesses/apollo/skills/design/data/motion/modal-center-arrive.md

@@ -1,18 +0,0 @@
----
-id: modal-center-arrive
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/review-animations/STANDARDS.md (physicality — modal exemption; duration table, modal row) + improve-animations/AUDIT.md §3; Material emphasized-decelerate agrees as secondary"
-license: open
-feel: "a centered scale-from-0.95-and-fade at ~250ms on a strong ease-out reads deliberate and important; scale(0) or a snap-open reads like it teleported in"
-aliases: { teleported: "scale(0) or an opacity-only entrance with no initial transform", "snap-open": "no interpolation at all — the dialog just appears" }
----
-Dialogs and confirmation modals only — the one place `transform-origin: center` is correct, exempt from the trigger-anchoring rule. Entrance starts from scale 0.9-0.97 + opacity 0, never scale(0); the longer end of the budget is the modal's prominence — don't reuse it on tooltips or dropdowns.
-
-```yaml values
-duration: 250ms
-easing: cubic-bezier(0.23, 1, 0.32, 1)
-element-class: overlay
-interruptible: n/a
-budget: { overlay: 200-500ms }
-```

package/data/harnesses/apollo/skills/design/data/motion/number-ticker-roll.md

@@ -1,18 +0,0 @@
----
-id: number-ticker-roll
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/animation-vocabulary/SKILL.md (Number ticker; Tabular numbers) + review-animations/STANDARDS.md (sub-300ms rule; interruptibility)"
-license: open
-feel: "digits rolling to the new value at ~250ms read like a live counter; an instant jump-cut reads like the number just glitched"
-aliases: { glitched: "value changes with no roll/interpolation at all", "tabular-drift": "digits set without tabular figures, so the number visibly shifts width as it rolls", janky: "keyframe roll restarting from zero on each rapid update instead of retargeting" }
----
-Live counters, stat tiles, KPI cards, running totals — system-driven value changes, so the sub-300ms UI rule applies. Requires tabular figures so surrounding layout doesn't shift while digits roll. Rapid successive updates retarget the roll (transition, not keyframes), never restart from zero.
-
-```yaml values
-duration: 250ms
-easing: cubic-bezier(0.23, 1, 0.32, 1)
-element-class: control
-interruptible: required
-budget: { control: 200-300ms }
-```

package/data/harnesses/apollo/skills/design/data/motion/onboarding-bounce.md

@@ -1,18 +0,0 @@
----
-id: onboarding-bounce
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/review-animations/STANDARDS.md (springs — bounce 0.1-0.3, reserved for playful moments) + improve-animations/AUDIT.md §1 (rare/first-time → delight)"
-license: open
-feel: "a springy overshoot on a first-run welcome card reads warm and considered; the same bounce on anything seen daily reads gimmicky"
-aliases: { gimmicky: "spring bounce applied outside a rare/first-time identity moment — the same overshoot on a daily-use control", overdone: "bounce pushed above the 0.1-0.3 range into cartoonish territory" }
----
-Onboarding welcome screens, first-run tours — rare, high-emotion, seen once or a handful of times ever. The scoped exception where springy/physical vocabulary is admissible; it does not license bounce anywhere else.
-
-```yaml values
-duration: 500ms (spring; no fixed duration)
-easing: spring(bounce 0.3)
-element-class: identity-moment
-interruptible: n/a
-budget: { identity-moment: 400-600ms }
-```

package/data/harnesses/apollo/skills/design/data/motion/panel-slide.md

@@ -1,18 +0,0 @@
----
-id: panel-slide
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/review-animations/STANDARDS.md (--ease-drawer curve; duration table, modal/drawer row) + apple-design/SKILL.md §3 (interruption; velocity continuity)"
-license: open
-feel: "a 260ms slide on the iOS drawer curve reads native and grabbable; a keyframe slide that restarts on repeated taps reads broken"
-aliases: { broken: "keyframes restarting from zero instead of a transition/spring retargeting mid-slide", floaty: "duration pushed past the panel budget" }
----
-Side panels, drawers, bottom sheets. Must be interruptible and carry velocity on a gesture-driven close — grabbing a closing panel should follow the pointer, not fight it.
-
-```yaml values
-duration: 260ms
-easing: cubic-bezier(0.32, 0.72, 0, 1)
-element-class: panel
-interruptible: required
-budget: { panel: 200-500ms }
-```

package/data/harnesses/apollo/skills/design/data/motion/press-feedback.md

@@ -1,18 +0,0 @@
----
-id: press-feedback
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/review-animations/STANDARDS.md (physicality — press feedback; duration table) + improve-animations/AUDIT.md §2-3"
-license: open
-feel: "160ms scale-to-0.97 on pointer-down reads tactile and instant; slower or absent reads dead under the finger"
-aliases: { dead: "no press feedback at all, or duration pushed above the 160ms control budget", mushy: "an easing softer than ease-out, delaying the visible response" }
----
-Any pressable control — buttons, icon buttons, tap-target rows. The canon prescribes this on every pressable element while forbidding animation on keyboard-initiated actions: it fires on pointer-down only (a tap affordance, not a state transition), and keyboard activation (Enter/Space) never replays it.
-
-```yaml values
-duration: 160ms
-easing: cubic-bezier(0.23, 1, 0.32, 1)
-element-class: control
-interruptible: n/a
-budget: { control: 100-160ms }
-```

package/data/harnesses/apollo/skills/design/data/motion/progress-linear-loop.md

@@ -1,18 +0,0 @@
----
-id: progress-linear-loop
-class: a6
-sourced: 2026-07-12
-source: "emilkowalski.skills/skills/review-animations/STANDARDS.md (easing decision order — constant motion → linear) + animation-vocabulary/SKILL.md (Loop; Marquee); Material's linear token agrees as secondary"
-license: open
-feel: "a constant linear sweep reads honest about an unknown wait; any easing that speeds up and slows down reads like a fake promise"
-aliases: { "fake-promise": "eased (non-linear) motion on an indeterminate progress indicator, implying a completion point that doesn't exist", twitchy: "loop cycle faster than ~1s, reading anxious rather than calm" }
----
-Indeterminate progress bars, loading spinners, marquees. Constant-speed motion only — the one place `linear` is correct and `ease` curves are wrong.
-
-```yaml values
-duration: 1200ms (loop)
-easing: linear
-element-class: control
-interruptible: n/a
-budget: { control: 1000-1500ms }
-```