AI turns documents or topics into real, native PowerPoint decks—with native shapes, transitions and animations, data-backed charts and tables on demand, audio narration from speaker notes, and support for your own .pptx templates. · by Hugo He
docs(patterns): absorb the full subtitle corpus into image, shape and animation refs
hugohe30b364752e99a
diff --git a/skills/ppt-master/references/animations.md b/skills/ppt-master/references/animations.mdindex f86292d0..0828e519 100644--- a/skills/ppt-master/references/animations.md+++ b/skills/ppt-master/references/animations.md@@ -162,6 +162,30 @@ Flags:**Hard rule — no silent downgrade**: an unknown transition effect, unsupported Effect Option, or invalid/non-finite duration fails export. It is never replaced by `fade`. Recorded narration keeps the resolved visual transition; `-t none --recorded-narration ...` writes narration-driven advance timing without restoring a visual effect.+### 3.1 Morph — author an action as the difference between two pages++Morph tweens objects it can match across consecutive slides. That makes it a general mechanism, not just a transition: **any continuous action can be authored as two static pages plus `-t morph`**, with no keyframe timeline anywhere. Duplicate the page, change one property on one object, and PowerPoint interpolates the rest.++| Change between the two pages | Reads as |+|---|---|+| Object sits off-canvas, then on-canvas | Slide-in, drawer pull, card extending |+| Object rotates | Flip, turn, hinge |+| Image container scales up | Camera push-in |+| Scrim opacity drops, or a cut contour grows | Progressive reveal |+| Same wide image at two `x` offsets | Camera pan (see image-layout-patterns `#87`) |++Chain three or more pages to build a sequence — extend, hold, retract — where each page is still an ordinary editable slide.++**Matching is by object identity, and it is strict**: keep the same image filename, the same group `id`, and container dimensions that do not change between the pages. Rename the file or resize the frame and morph silently degrades to a cross-fade with none of the motion. This is the most common reason a morph sequence "does nothing".++**Give text somewhere to come from.** Morph tweens objects present on both pages; text that only exists on the second page can only fade in. The standard fix, used in essentially every morph-driven deck: place the *next* page's copy on the current page just outside the canvas (below), and the *previous* page's copy just outside the opposite edge (above). Each block then slides through the frame instead of blinking, and the deck reads as one continuous surface being scrolled. Objects parked outside the canvas are not rendered but must still exist on both pages with the same identity.++**When morph refuses to match**: PowerPoint pairs objects of the same kind first, so two different shape types, or a shape and a picture, will cross-fade instead of tweening. Authored decks force the pairing by giving both objects an identical custom shape name. The exporter does read a per-object name — `data-pptx-shape-name` on the wrapper — but that attribute is currently specified as **importer metadata** for mirror/preserve packages ([`svg-effects.md`](./svg-effects.md) §6.6), not as an authoring control for generated pages. Until that contract is widened, keep morph pairs the same object kind with matching geometry and `id`, and do not introduce the attribute on generated pages to force a match.++**Not supported — Slide Zoom / Summary Zoom.** Click-to-jump navigation built on PowerPoint's Zoom objects (the "click a portrait, zoom into that section" pattern) has no exporter path. Build click-driven navigation with `trigger_shape` on ordinary object animations instead, or with plain hyperlinks.++**No 3D**: perspective rotation, extrusion, and shear are outside the SVG contract — `skewX` / `skewY` and shear matrices fail closed ([`svg-effects.md`](./svg-effects.md) §6.8). Build the same impression with 2D means — offset, scale, overlap, and per-facet lightness (image-layout-patterns `#91`) — rather than attempting a 3D tilt.+---## 4. Per-Element Animations@@ -233,6 +257,29 @@ motion but preserves narration audio and recorded advance timing.> Note: `--recorded-narration` rejects `on-click` and `trigger_shape`; use its default `narration_animations.json`, pass `--animation-config animations.json` for the canonical presentation animation, or pass `--no-animations`.+### 4.1 Slow ambient motion — the page that breathes++Most object animation exists to reveal content on click. There is a second, quieter use: **one long, slow motion on the background that never waits for a click**, so a static page stops looking frozen. It is the highest-impact-per-effort motion available and it costs one sidecar entry.++The recipe: a `path_left` (or `path_right`) motion on the **background image group only**, `with-previous` so it starts unprompted, and a duration of **4–10 s** — an order of magnitude longer than the reveal default. Set the travel distance so the image is still fully covering the canvas at both ends; a background that drifts past its own edge exposes the slide beneath it.++It pairs naturally with a fixed foreground: with image-layout-patterns `#90`, the scrim and its cut contour stay locked while the world moves behind the cuts, which reads as looking through windows rather than as a sliding photo. The same logic applies to `#82` and `#12`.++**Restraint is the whole technique**: one moving object per page, background only, never body copy or data. Two ambient motions on one page cancel each other out and the page reads as unstable.++### 4.2 Recurring recipes++Four combinations that recur constantly in authored decks. Each is built from+mechanisms already defined above — none needs a new capability.++**Carousel** (morph, §3.1) — hold a fixed row of card frames and rotate the *content* through them: on each page every image advances one position, so the card at centre changes while the frames stay put. Morph then slides the images between frames and the row appears to scroll. Requires identical frame geometry and `id`s on every page; only the image assignments change. Scales to any number of images with one page each.++**Odometer / counting numerals** (morph or motion path) — build a vertical strip of digits 0–9 and show one through a fixed window: a masked opening, or a background-filled rectangle above and below ([`image-layout-patterns.md`](./image-layout-patterns.md) `#95`). Shift the strip so the target digit lands in the window, then either morph between two pages or run a `path_up` motion on the strip. Give each digit column a 0.1 s stagger so they settle in sequence rather than in lockstep.++**Parallax depth** (morph) — move a background layer a *short* distance and a foreground layer a longer one between two pages. The differing travel is read as depth. Keep both layers' z-order identical on both pages; a layer that changes stacking between pages breaks the tween and the transition jumps.++**Flip-card / click-to-reveal** (`trigger_shape`, §4) — pair a face group and a back group at the same position, give the face an exit and the back an entrance, and set the back's `trigger_shape` to the face's id. Clicking the face plays both. This is the supported route for click-driven interaction; PowerPoint's Zoom objects are not (§3.1).+---## 5. Anchor Logic — Top-Level `<g id="...">`diff --git a/skills/ppt-master/references/image-layout-patterns.md b/skills/ppt-master/references/image-layout-patterns.mdindex cdf296f9..1a60a0f4 100644--- a/skills/ppt-master/references/image-layout-patterns.md+++ b/skills/ppt-master/references/image-layout-patterns.md@@ -4,7 +4,7 @@ A vocabulary registry of ways images can be placed on a slide. The point of thisEvery entry has a name plus a short technical hint. Common techniques get a single line. Less obvious or easily forgotten techniques get a short paragraph — not a full tutorial, but enough that a model unfamiliar with the project can implement it without guessing. This is a registry, not a teaching document; no use-case prescriptions, no decision tables.-> **Numbers are stable identifiers, not sequence.** The file is split into **Part 1 — Primary Structures** (#1–#19, #38–#56, #73–#81, #88) and **Part 2 — Modifier Layers** (#20–#37, #57–#72, #82–#87, #89–#91). Numbers jump within each Part because Primary structures were grouped first; existing references to `#38`, `#48`, etc. anywhere in the project still resolve correctly.+> **Numbers are stable identifiers, not sequence.** The file is split into **Part 1 — Primary Structures** (#1–#19, #38–#56, #73–#81, #88, #92–#94) and **Part 2 — Modifier Layers** (#20–#37, #57–#72, #82–#87, #89–#91, #95–#99). Numbers jump within each Part because Primary structures were grouped first; existing references to `#38`, `#48`, etc. anywhere in the project still resolve correctly.---@@ -44,6 +44,18 @@ Pick one or more of these as the page's bones. Cross-primary combinations are en9. **3×3 grid with central image** — nine cells; center cell holds the image, the other 8 hold text blocks, color swatches, or small data widgets.+93. **Containers arrayed along a curve (fan, arc, ring)** — N image containers distributed along an arc or wave, each rotated to sit square to the curve at its own position. Reads as motion and hierarchy at once, and it is the backbone of fan spreads, ring layouts, dial/roulette pages, and arched photo rows.++ **Geometry** — place container `i` of `n` on a circle of radius `r` about `(cx, cy)`:+ ```+ θᵢ = θ_start + i × (θ_span / (n − 1))+ xᵢ = cx + r·cos(θᵢ) yᵢ = cy + r·sin(θᵢ)+ rotationᵢ = θᵢ + 90° (tangent-aligned; drop this for upright containers)+ ```+ Use `transform="rotate(rotationᵢ xᵢ yᵢ)"` on each container group. A `θ_span` of 60–120° reads as a fan; 360° with `θ_start = -90°` gives an evenly spaced ring. For a wave instead of an arc, sample the wave's own path and use its local tangent as the rotation.++ **Two things to get right**: keep radius and angular step *constant* — an eyeballed fan reads as a mistake, not a flourish; and when containers are tangent-aligned, images inside must not inherit the rotation blindly (a sideways face is the failure mode). Counter-rotate the image inside its container, or keep the containers upright and let only their positions follow the curve.+10. **Centered image with radial callouts pointing outward** — image (often circular via `clipPath`) at canvas center; multiple `<line>` leader lines + small `<circle>` endpoints + offset text labels in surrounding space.11. **Diagonal split with directional gradient (not hard polygon cut)** — full-bleed `<image>` + overlay `<rect fill="url(#grad)">` whose gradient axis runs along the diagonal, plus a `<line>` to make the divider read. Do NOT hard-clip: polygon cuts give stair-stepped edges on text panels.@@ -88,6 +100,12 @@ This is the family that opens up the largest design space and the one AI is most## Multi-Image Compositions+94. **Embracing arc row (2D substitute for a 3D perspective wall)** — a row of images or cards where the centre element is largest and each step outward shrinks and drops, so the tops trace an arc and the row appears to curve toward the viewer. This is what PowerPoint decks build with 3D rotation (perspective left / right, X-axis 330° / 30°) for logo walls, certificate rows, and photo shelves — and it is reproducible in 2D, which matters because 3D transforms are outside the SVG contract ([`svg-effects.md`](./svg-effects.md) §6.8).++ **Construction**: for element `k` steps from the centre, apply `scale = 0.88ᵏ` and offset `y` downward so every element's *top* edge lands on one shallow arc; keep the horizontal step constant. Mirror the sequence left and right of the centre. Add a soft ground shadow or a reflection fading downward to seat the row. Bottom-aligning instead of top-arcing gives the flatter "shelf" variant.++ The depth cue is entirely **scale + vertical offset + consistent light**; do not reach for skew or a fake 3D tilt, which fail closed on export. Three to seven elements is the working range — beyond that the outermost ones shrink into illegibility.+47. **Small multiples — 3–6 same-kind images in an evenly spaced row** — identical containers, identical caption blocks (title + one line). Not a generic grid: the identical framing *is* the message, because readers compare across panels only when the structure is constant.48. **Side-by-side comparison (before/after, A/B, then/now)** — two `<image>` of equal size in 50/50 split with thin divider `<line>` and "before" / "after" labels.@@ -98,6 +116,19 @@ This is the family that opens up the largest design space and the one AI is most51. **Mosaic** — irregular tile sizes packed together with or without thin gaps; each image clipped to its tile's rect.+92. **Split tiling — one parent shape cut into interlocking cells** — the most-used construction in real image-heavy decks, and the counterpart to #82. Take one parent shape (circle, annulus, rounded rect, trapezoid, wave band), lay cutting lines across it (long bars, evenly distributed or fanned at different angles), and split it into cells. Each cell then holds a *different* image. Because every cell comes from one parent, the edges interlock exactly — no gaps, no overlaps, and the group still reads as one object.++ | Parent + cutters | Result |+ |---|---|+ | Circle + 2 crossed bars | Quadrant wheel |+ | Annulus + radial bars | Ring segments |+ | Wave band + vertical bars | Rhythmic strip |+ | Trapezoid + slanted bars | Perspective row |++ **Authoring**: compute each cell's contour and write it as its own `<path>` clip — the geometry is deterministic, so derive the cells rather than eyeballing them. `shape_boolean_svg.py fragment` returns exactly these interlocking regions as separately addressable paths. Give every cell the same stroke (2px, background color) so the cuts read as designed seams.++ **Choosing between #92 and #82**: same construction, opposite content rule. One image across all cells (#82) says "these fragments are one thing"; a different image per cell (#92) says "these are peers, cut from one frame". Mixing them destroys both readings. Distinct from #50 / #51, where cells are independent rectangles that never shared a parent.+52–53. **Filmstrip / stack** — a sequence of `<image>` with thin consistent gaps: horizontal, equal height and varying widths (**#52**), or vertical, aligned by width with shared annotations down one side (**#53**).54. **Overlapping image stack** — `<image>` elements with overlapping `x/y` positions; each subsequent one in front (z-order by document order); often combined with slight rotation for layered photo-print look.@@ -182,11 +213,44 @@ Stack any of these freely on top of a Primary structure. Multiple Modifiers per30–31. **Flat overlay wash** — one `<rect fill-opacity>` over the image: neutral `#000` / `#fff` around 0.4 for uniform darkening or lightening, the simplest scrim there is (**#30**), or a deck color at 0.15–0.25 to pull a foreign-looking photo toward the palette without regenerating it (**#31**).+> **Sample the scrim color from the photo itself.** For any gradient scrim over an image (#27, #29, #31, #32, #90), take the solid end's hex from a dominant color *in that image* rather than defaulting to black or a deck color, and slide the gradient stop until the seam between scrim and photo disappears. A black scrim over a warm photo announces itself as a rectangle; a scrim in the photo's own shadow tone reads as part of the picture. This one substitution is the difference between a page that looks masked and one that looks composed.++98. **Grid scrim with per-cell opacity** — instead of one flat or gradient scrim, cover the image with a grid of adjacent rectangles and give each cell a *slightly different* opacity (say 10–40 %, varied irregularly). The photo shows through unevenly, so the overlay reads as texture — panelled glass, a pixel field, a contact sheet — rather than as a sheet of paint. Text sits on the denser cells.++ Keep the variation small and non-repeating: a regular light/dark alternation reads as a checkerboard, and a wide spread reads as broken rendering. Butt the cells exactly (no gaps, no strokes) so the grid is felt rather than drawn. Distinct from #50 / #88, where every cell holds its own image; here one image lies beneath one grid of glass.++99. **Selective desaturation — colour only where it matters** — the whole image is muted while one subject stays in full colour, which fixes the focus of a busy photo without cropping it. Two registered copies: a desaturated (and usually darkened) version filling the frame, and the colour original clipped to just the subject region, sitting exactly on top.++ **Both copies are baked assets** — there is no runtime colour filter on the native route ([`svg-effects.md`](./svg-effects.md) §6.12), so produce the desaturated file with a one-line Pillow `ImageEnhance.Color(img).enhance(0)` pass rather than reaching for `feColorMatrix`. Clip the colour copy along a real edge in the picture (the subject's own contour, per #96) — a rectangular colour patch over a desaturated field reads as an accident.+32. **Multi-stop scrim with hue shift** — three-or-more-stop `<linearGradient>` where stops are different colors (e.g. dark navy → transparent → warm orange). This re-grades the image's color world without regenerating — particularly useful when an AI image came back with the right composition but wrong color temperature.-90. **Full-canvas scrim with a shape cut out of it (chapter-page formula)** — a full-slide `<path>` whose outer contour is the canvas and whose inner subpath is a wave, arc, ribbon, or oversized numeral, cut using the opposite-winding rule from #83. Fill the scrim with a gradient whose stops vary `stop-opacity` (e.g. `1 → 0.8 → 0`) rather than color, so the underlying image emerges progressively across the page instead of showing through one hard hole. Add a 1–2px light stroke on the cut edge to keep the boundary crisp. This is the most reliable chapter/divider formula in the catalog: one image, one scrim, one numeral.+90. **Full-canvas scrim with shapes cut out of it (the cover / divider formula)** — the single highest-yield formula in this catalog, and the one real decks reuse most: a full-slide `<path>` whose outer contour is the canvas and whose inner subpath(s) are cut out using the opposite-winding rule from #83, laid over a full-bleed image. The scrim mutes the photo everywhere except through the cuts, so one ordinary image becomes a designed page. Three elements total: image, scrim, title.++ **The cut contour** — any of these, all authored as reversed inner subpaths in the same `<path>`:- **Numeral / lettering caveat**: cutting *text* out of the scrim requires the glyph as a `<path>` outline, which is not something to author by hand — least of all for CJK. For a chapter number, either set the numeral as ordinary `<text>` on top of the scrim (nearly as strong, fully editable), or pre-render the knocked-out numeral as an RGBA PNG (#68). Do not attempt to approximate glyph outlines.+ | Contour | Reads as |+ |---|---|+ | Wave, arc, ribbon (one soft curve across the page) | Editorial banner / horizon |+ | Freehand closed curve (irregular, hand-drawn) | Organic torn-paper window |+ | An array of hexagons / trapezoids / circles | Rhythmic screen, a window wall |+ | Oversized numeral or letterform | Chapter marker (see caveat) |++ An array of cuts is just several reversed subpaths in the same `d` — the same construction as #82, except here the image shows *through* the holes rather than being clipped *into* the shapes.++ **Paint the scrim** with either a flat light fill at 0.15–0.25 opacity (white over a photo is the reliable default) or, for a directional reveal, a gradient that varies `stop-opacity` rather than color (`1 → 0.8 → 0`), so the image emerges progressively instead of through one hard boundary. Add a 1–2px stroke in the same light color on the cut edge to keep it crisp.++ **Edge thickness**: to make the cut read as a physical opening, apply `feDropShadow` with `dx="0" dy="0"` and a small `stdDeviation` to the scrim path. Per [`svg-effects.md`](./svg-effects.md) §6.4 a zero-offset shadow is classified and exported as a **glow**, not a shadow — so use an accent or light color; black will read as diffuse haze rather than an edge. Never apply it to the `<image>` itself (#36).++ **Numeral / lettering caveat**: cutting *text* out of the scrim needs the glyph as a `<path>` outline, which is not something to author by hand — least of all for CJK. Set the numeral as ordinary `<text>` over the scrim (nearly as strong, fully editable), or pre-render a knocked-out numeral as an RGBA PNG (#68). Do not approximate glyph outlines.++ **Motion pairing**: the scrim stays fixed while the image beneath drifts slowly (a 4–10s linear path, left, starting with the previous animation) — the cuts then behave like windows onto a moving world. That is an animation-stage decision, not page design; see [`animations.md`](./animations.md). It pairs with this pattern more often than with any other.++97. **Frosted-glass panel over the photo** — a legibility panel that is neither a flat scrim (#30) nor a full blur: a region of the image itself, blurred and lightened, sitting under the text while the rest of the photo stays sharp. It keeps the photo's color and composition visible through the panel, which a solid scrim destroys.++ **Build it from a baked asset** — runtime blur does not survive native export ([#34](#), [`svg-effects.md`](./svg-effects.md) §6.12). Produce a blurred copy of the source (a Pillow `GaussianBlur` at a large radius, plus a brightness lift), then place that copy clipped to the panel contour and registered to the same position as the base image, so the blur lines up exactly with what is behind it. Add a thin light stroke and, if the style wants it, a slight lightening overlay.++ The panel must stay in register with the base photo; a frosted panel showing a *different* part of the scene is the classic tell. Pair with #95 when the panel should also carry a floating edge.33. **Spotlight mask — clear region surrounded by darkness** — cover the canvas with `<rect>` filled by a `<radialGradient>` whose inner stop is fully transparent and outer stop is opaque dark. Reads as a flashlight beam on the focal area. Use sparingly — it kills everything outside the spotlight.@@ -230,6 +294,23 @@ Stack any of these freely on top of a Primary structure. Multiple Modifiers per72. **Image-to-image transition / merge** — two `<image>` elements with overlapping regions, one or both with gradient masks (from group C) creating a soft blend between them.+95. **Shape filled with the page background itself** — the most-used trick in real decks and the one that has no obvious SVG name. A shape is painted not with a color but with *the page's own background, sampled at the shape's own position*, so it becomes invisible against the page while still being a real object that can be moved, animated, or given an edge.++ **SVG form**: give the shape the same `<image>` as the page background, positioned in root coordinates exactly as the background is, and clip it to the shape contour (§1.2). Because the fill stays registered to the page rather than to the shape, the object reads as a hole in whatever is above it.++ Three things it buys you, all of which otherwise require a second asset:+ - **A cut that keeps the scene continuous** — the shape "removes" a foreground panel and shows the background through it, with no seam even over a photo or gradient.+ - **Invisible objects that still animate** — a background-filled bar can wipe, slide, or morph across the page to reveal or conceal content, while never being visible itself.+ - **Edge-only forms** — the shape disappears but its stroke, glow, or shadow remains, giving a floating outline that appears cut into the page.++ Distinct from #83 (a panel with a real hole) and #90 (a scrim with cuts): those remove paint, this one *impersonates* the background. Reach for it when the thing above must stay a solid object.++96. **Cutout subject re-laid over its own photo** — the mechanism behind every "subject escapes the frame" page (#85), and worth stating on its own because it is a two-asset technique: keep the original photo as the background layer, and place a background-removed PNG of its subject on top, in perfect register.++ Once the subject exists as a free-floating layer, it can overlap anything drawn between the two copies: a title the subject stands in front of, a color panel it steps out of, a shape frame it breaks through, a grid line it crosses. The base photo can be tinted, desaturated, blurred (baked), or scrimmed as hard as the layout needs, because the sharp subject on top is what the eye reads.++ Register is everything — the cutout must sit exactly where the subject sits in the base image; a few px of drift reads as a printing error. Keep the cutout's own edge clean rather than adding a stroke, unless the design calls for the sticker look of #63.+89. **Same image twice — sharp cutout over a receded full-bleed copy** — the single best answer to "the photo is too narrow / too short for this canvas, and stretching distorts the subject". Reference the same file twice: the bottom copy fills the whole canvas (or panel) and is pushed back; the top copy is clipped to a shape (#82, #24, a slanted band, a folded contour) at native proportions and stays sharp. The subject reads at full fidelity while the background extends the frame to any aspect ratio — no stretching, no letterbox bars, no second asset.**Recede the bottom copy with what survives export**: a color-tinted or darkened overlay `<rect>` (#30 / #31) at 0.5–0.8, or a desaturated / lowered-brightness variant of the file. **Blur does not survive** — per #34 the native route does not preserve a blurred-image backdrop, so if the design depends on blur it must be baked into a second image file (a one-line Pillow `GaussianBlur` pass over the original is enough); never rely on a filter at export time. Keep both copies in register — same center, same crop logic — or the trick reads as two unrelated photos.@@ -279,6 +360,8 @@ Combine freely. The "AI-default" failure mode is the opposite: defaulting to bar**When the supplied image does not fit the canvas**, the answer is #89, not stretching and not letterboxing: the same file placed twice, receded behind and sharp in front. This is the single most common image problem in real decks and it has a purely geometric solution.+**When one ordinary photo has to carry a cover or divider**, reach for #90 before anything else. A full-canvas scrim with shapes cut out of it turns any stock image into a designed page using three elements, and its cut contour (curve, freehand shape, tessellated array, numeral) is where the page's character comes from. It is the most reused formula in real image-led decks.+**Skip-detection signal** — if every page's `Layout pattern` column resolves to bare #2 / #3 / #5 / #6 with no Modifier ids, the catalog was not consulted. Re-read and reconsider.**Cross-page through-line (recurring motif).** The patterns above are per-page, but a deck reads as *designed* when one illustration motif family recurs across pages—a cover anchor, section dividers repeating the motif (`#75`), and small `#63` spots threaded through the body. Keep one family (shared rendering / locked deck colors / subject world), vary scale and placement, and never turn recurrence into a quota.diff --git a/skills/ppt-master/references/native-shape-authoring.md b/skills/ppt-master/references/native-shape-authoring.mdindex 73f66db6..8fa02e69 100644--- a/skills/ppt-master/references/native-shape-authoring.md+++ b/skills/ppt-master/references/native-shape-authoring.md@@ -228,3 +228,84 @@ stores the materialized freeform geometry, not replayable operation history.page. In one normal `apply_patch` edit, remove every selected operand and insertevery returned path at the SVG root in the primary operand's z-order. Fragmentpaths remain separate shapes; do not wrap them to claim one structural atom.++---++## 7. Shape-Only Modelling Techniques++Applies to any page built from shapes, **with or without images** — a text-only,+data-only, or icon-only deck reaches these the same way. Each technique below is+plain geometry plus gradient paint, so all of it survives native export.++### 7.1 Alternating light/dark gradient = dimensional form++The single highest-yield shape technique. A cylinder, metallic band, dimensional+numeral, or curved panel is produced by one gradient whose stops **alternate+light and dark** across the shape — light · dark · light for a three-stop ramp,+or light · dark · light · dark · light for a five-stop one. The alternation+imitates a curved surface catching light twice; a plain two-stop ramp always+reads flat no matter how strong the contrast.++Keep every stop on one hue and vary only lightness, hold one light direction for+the whole page, and remove strokes so adjacent facets meet cleanly. For a+cylinder, apply the alternating ramp across the body and cap it with an ellipse+carrying its own shallower ramp. This is the shape-level twin of+[`image-layout-patterns.md`](./image-layout-patterns.md) `#91`, which applies the+same idea across separate facets of a folded form.++### 7.2 Reflection without a reflection effect++Native reflection is `Bake-required` ([`svg-effects.md`](./svg-effects.md) §6.12),+so build it from geometry instead:++1. Duplicate the object and flip it with `transform="translate(0, 2·y_bottom) scale(1, -1)"`.+2. Keep only the top **10–25 %** of the flipped copy — that is all a reflection+ ever shows.+3. Lay a rectangle over it filled with a gradient running from fully transparent+ at the object's base to the page background color at the cut line, so the+ copy dissolves into the page.+4. Drop the whole reflection to roughly **60–70 %** opacity.++Seat rows of certificates, product shots, logo tiles, and cylinders this way. Do+not add a blur — it will not survive export, and a short gradient fade already+reads correctly at slide scale.++### 7.3 Fragment as a modelling tool, not just a boolean++`fragment` (§6) is the fastest way to build layered diagrams from one silhouette:+lay evenly distributed bars across a triangle and fragment it into pyramid tiers;+cross a circle with two bars for a quadrant wheel; slice an annulus radially for+ring segments. Every piece inherits the parent contour, so the assembly stays+perfectly registered — impossible to achieve by drawing the tiers separately.++Distribute the cutting bars with a constant step before fragmenting; uneven tiers+read as a mistake rather than a hierarchy. Paint the resulting pieces with one+gradient family per §7.1 so the stack reads as a single solid.++### 7.4 Soft edges without the soft-edge effect++Feathered edges are `Bake-required` ([`svg-effects.md`](./svg-effects.md) §6.12),+but the four jobs they normally do are all reachable with gradients:++| Intent | Build instead |+|---|---|+| Contact shadow under an object | Ellipse filled with a `radialGradient` from dark-transparent at the centre to fully transparent at the rim |+| Spotlight / stage pool | Cone or ellipse filled with a gradient fading to transparent at its far end, at low opacity over the scene |+| Object dissolving into the page | Overlay a rectangle whose gradient runs from transparent to the exact page background hex |+| Hiding an object while keeping it live | Full transparency, or a background-registered fill ([`image-layout-patterns.md`](./image-layout-patterns.md) `#95`) |++A radial or linear alpha ramp reads the same as a feathered edge at slide scale+and, unlike a filter, exports intact. Never approximate a soft edge with a stack+of stroked outlines — the banding is visible on projection.++### 7.5 Ground plane and staging++An object floating in empty canvas looks pasted on. Give it a surface: a wide+shallow ellipse or trapezoid beneath it, filled with a gradient that fades to the+background at its edges, optionally with a soft dark ellipse directly under the+object as contact shadow. A trapezoid narrowing away from the viewer reads as a+receding floor; a cylinder or slab reads as a pedestal.++Keep the plane low-contrast — it is staging, not content. This is what makes+certificate rows, product hero shots, and trophy/award pages look composed+rather than floating, and it costs two shapes.diff --git a/skills/ppt-master/workflows/generate-pptx.md b/skills/ppt-master/workflows/generate-pptx.mdindex ae57a454..b51bb856 100644--- a/skills/ppt-master/workflows/generate-pptx.md+++ b/skills/ppt-master/workflows/generate-pptx.md@@ -357,7 +357,7 @@ Read references/visual-styles/<resolved-id>.md # one preset id, or each `visu| `spec_lock.md images` or §VIII contains at least one image/formula row, or an active template carries bundled images | `executor-image.md` + `image-layout-patterns.md` + `image-layout-spec.md` + `svg-image-embedding.md` || At least one placed image has `Status: Sourced` | `executor-web-image.md` after the image branch || The locked style/current page calls for noncanonical or alpha paint, dash/cap/join, tracking/decoration/outline, gradient/filter/glow/shadow, path/transform/clipping, or another constructed effect | `svg-effects.md` before authoring that value or effect |-| A page calls for a literal PowerPoint stock shape or an explicit Merge Shapes operation | `native-shape-authoring.md` before selecting or materializing that geometry |+| A page calls for a literal PowerPoint stock shape, an explicit Merge Shapes operation, or shape-built dimensional form (cylinder/pedestal, layered diagram, reflection, ground plane) | `native-shape-authoring.md` before selecting or materializing that geometry. This trigger is image-independent: a text-only, data-only, or icon-only page reaches it the same way || All SVG pages and SVG quality gates are complete | `executor-notes.md` before generating speaker notes |No branch is loaded by analogy. Evaluate these triggers from `spec_lock.md`, §VII/§VIII, the selected style, and the current page plan.