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
refactor(image-types): narrow type to "internal geometric skeleton"; fold hero/background/portrait/typography into page_role primitives
hugohe3a755a11569a8
diff --git a/docs/roadmap.md b/docs/roadmap.mdindex bf04950d..4a177b98 100644--- a/docs/roadmap.md+++ b/docs/roadmap.md@@ -35,6 +35,7 @@ The past two months' structural capability growth. Single flags / incremental po- **AI image `hero_page` dual-track** — local insert + full-canvas hero image coexist- **Brand identity preset subsystem** ([`workflows/create-brand.md`](../skills/ppt-master/workflows/create-brand.md)) — extract and reuse brand palette / typography / logo / voice- **Visual self-review workflow** ([`workflows/visual-review.md`](../skills/ppt-master/workflows/visual-review.md)) — rubric-based per-page check of AI-generated SVGs+- **AI image: Type concept boundary clarification** — Type is now narrowed to "the internal geometric skeleton of a local infographic block" (11 real skeletons); the four pseudo-types (hero / background / portrait / typography) fold back into `page_role: hero_page` plus four composition primitives (single-subject / portrait / typographic / atmospheric); hero_page text layering rule (visual keywords embedded, editable text via SVG overlay)---@@ -42,15 +43,7 @@ The past two months' structural capability growth. Single flags / incremental poStrong rationale, work in progress.-### 1. AI image: Type de-bloat--There are currently 15 types (hero / background / portrait / typography / matrix / cycle / funnel / pyramid / map / scene / infographic / metaphor / flowchart / comparison / framework, etc.) mixing "internal composition" and "PPT role" — hard for the AI to pick correctly.--Plan: collapse to ~6 core types (infographic / scene / flowchart / comparison / framework / timeline); hero / background / portrait / typography move into the `page_role` dimension; matrix / cycle / funnel / pyramid become framework sub-types; map becomes a scene sub-type.--Large blast radius (13 type files + manifest schema), will land in stages.--### 2. Capability-backing example decks (Style demo P0)+### 1. Capability-backing example decks (Style demo P0)As the community enters its growth phase, example decks shift from "show visual contrast" to "prove the capability bar" — making it obvious why Gamma / Manus-style tools can't replace PPT Master.diff --git a/docs/zh/roadmap.md b/docs/zh/roadmap.mdindex c30a68f5..aba056bd 100644--- a/docs/zh/roadmap.md+++ b/docs/zh/roadmap.md@@ -35,6 +35,7 @@- **AI 图 `hero_page` 双档** — 局部插图 + 整页主角图共存- **品牌身份预设子系统**([`workflows/create-brand.md`](../../skills/ppt-master/workflows/create-brand.md)) — 提取并复用品牌色板 / 字体 / Logo / 语调- **视觉自检工作流**([`workflows/visual-review.md`](../../skills/ppt-master/workflows/visual-review.md)) — 按 rubric 逐页自查 AI 生成的 SVG+- **AI 图 Type 概念边界澄清** — Type 收窄回「local 信息图的内部几何骨架」(11 个真骨架);原 4 个伪 type (hero/background/portrait/typography) 折回 `page_role: hero_page` + 4 条构图通则(single-subject / portrait / typographic / atmospheric);hero_page 文字分层规则(关键视觉词 embedded、可改文字走 SVG)---@@ -42,15 +43,7 @@立项依据强、近期会动手的方向。-### 1. AI 图 Type 反膨胀--现有 15 个 type(hero / background / portrait / typography / matrix / cycle / funnel / pyramid / map / scene / infographic / metaphor / flowchart / comparison / framework 等),其中一部分是「内部构图」描述,一部分是「PPT 用途」描述,两类混在一起 AI 不好选。--计划:折叠到约 6 个核心 type(infographic / scene / flowchart / comparison / framework / timeline);hero / background / portrait / typography 归入 `page_role` 维度;matrix / cycle / funnel / pyramid 作为 framework 子类;map 作为 scene 子类。--影响面较大(13 个 type 文件 + manifest schema),会分阶段做。--### 2. 能力背书型示例 deck(Style demo P0 三档)+### 1. 能力背书型示例 deck(Style demo P0 三档)社区进入增长期后,示例 deck 的作用从「展示视觉反差」切换到「证明能力门槛」—— 让用户一眼看懂为什么 Gamma / 美图 / Manus 这条线替代不了 PPT Master。diff --git a/skills/ppt-master/references/image-generator.md b/skills/ppt-master/references/image-generator.mdindex adc07609..3734a559 100644--- a/skills/ppt-master/references/image-generator.md+++ b/skills/ppt-master/references/image-generator.md@@ -46,7 +46,7 @@ Every AI image is described by three orthogonal dimensions. Lock them in this or|---|---|---|| **Rendering** | Visual style family (vector / sketch-notes / 3d-isometric / corporate-photo / …) | Once per deck — every AI image in the deck shares one rendering || **Palette** | How the deck's HEX colors are *used* (proportion + role + temperament). HEX values come from `design_spec.colors`, not from the palette | Once per deck |-| **Type** | What the image's internal composition looks like (background / hero / infographic / framework / comparison / timeline / scene / flowchart / typography) | Per image |+| **Type** | What the image's internal composition skeleton looks like — geometric layout of a local infographic block (infographic / flowchart / framework / matrix / cycle / funnel / pyramid / comparison / timeline / map / scene). Only applies to `page_role: local`; for `page_role: hero_page`, describe composition with §4.1 primitives instead of picking a type. | Per image |> **What rendering vs palette means**: rendering is *how the image is drawn* (line quality, texture, depth). Palette is *how colors are distributed and behave* (which color dominates, which is accent, what proportion). The HEX values come from Strategist; palette is the **usage contract** for those HEX values.@@ -114,7 +114,7 @@ Then `read_file` the **single resolved** rendering file and the **single resolveFor each `Acquire Via: ai` row in `design_spec.md §VIII`:-1. **Determine type** by matching the row's `Purpose` against types `_index.md` auto-selection table (cover background → `background`; product launch hero → `hero`; methodology visualization → `framework`; etc.) The narrative-shorthand `Type` column in §VIII (Background/Photography/Illustration/Diagram/Decorative) is a hint, not the type's final value — `Purpose` is authoritative for picking among the 9 internal-composition types.+1. **Determine type** — only when `page_role: local` (the image sits as a region block on an SVG page). Match the row's `Purpose` against the `_index.md` auto-selection table (methodology visualization → `framework`; process steps → `flowchart`; SWOT/Eisenhower → `matrix`; PDCA / flywheel → `cycle`; etc.). The narrative-shorthand `Genre` column in `design_spec.md §VIII` (Background / Photography / Illustration / Diagram / Decorative) is a hint, not the type's final value — `Purpose` is authoritative for picking among the 11 internal-composition types. **When `page_role: hero_page`, skip type selection** and describe composition directly using §4.1 primitives (single-subject / portrait / typographic / atmospheric).2. **Determine `text_policy`** — Strategist's value wins when set. Otherwise pick `none` or `embedded` based on whether in-image text serves the page. Long body / data / lists stay in SVG.3. **Determine `page_role`** — Strategist's value wins when set. Otherwise pick `local` or `hero_page` based on whether the image carries the page or sits inside one.4. `read_file references/image-type-templates/<type>.md` (only if not already read — types are commonly reused across images in one deck)@@ -157,6 +157,52 @@ Every assembled prompt follows this paragraph structure. **Write prose, not tagThis produces generic, model-average output. The model is not weighting your tags — write **one coherent visual scene** instead.+### 4.1 Hero-page composition primitives++When `page_role: hero_page` (the image is the page's main voice — cover, chapter divider, mood transition, signature stat, closing quote), the image's internal composition does not need its own structural `type` (matrix / cycle / framework etc. are for *local* infographic blocks). Instead, describe the composition directly in the prompt using one of the four primitives below.++**Primitive A — single dominant subject (product / object / concept hero)**++> One dominant subject occupying 60-70% of the canvas, positioned with intent (centered, rule-of-thirds offset, or slight left/right). Supporting context <30% of canvas weight. Generous negative space — at least 15% padding on the subject's "open" side. No second-place subject competing.++Use for: product reveal, concept introduction, chapter title visual, brand statement.++**Primitive B — single human subject (portrait)**++> One person, frontal or three-quarter turn, head + upper body. Subject occupies 50-65% of canvas height, centered or rule-of-thirds offset. Eyes at the upper-third horizontal line. Background neutral, minimal, or softly blurred. No competing foreground objects. At least 15% padding above the crown.++Use for: founder profile, speaker bio, testimonial page, executive intro. Pair with `rendering: corporate-photo` for photographic realism; otherwise the §5.2 simplified-figures rule applies.++**Primitive C — typographic hero (the text *is* the image)**++> The image's central content is one large text element — a short headline, big number, or single word — rendered as art, occupying 40-60% of canvas height. Minimal supporting visual (small icon, geometric anchor, accent line) at <25% weight. At least 20% padding around the text.++Use with `text_policy: embedded`. Must obey the §5.3 text-accuracy rules (English short-text and simple numbers reliable; CJK and long sentences fail — for those, switch to Primitive D + SVG overlay).++**Primitive D — atmospheric backdrop (no subject)**++> Atmospheric field with no dominant subject — gradients, subtle patterns, or restrained color blocks. Small geometric anchor optional, placed in a corner or along an edge, never centered. The center 60-70% of the canvas must stay calm to receive SVG title/text overlay.++Use for: cover background, chapter divider background, breathing-page background, any page where the SVG layer carries the words and the image only sets tone.++**Fewshot examples per primitive** (one each, deck-context placeholders intact):++> **A — 3d-isometric + tech-neon product reveal, text_policy: none, 600×600**+>+> 3D isometric illustration in true 30°/30°/30° projection. One dominant product-form subject — a stylized device or sleek tech object — occupies the center of the canvas at roughly 65% of the area. The subject is rendered in primary electric blue `#0EA5E9` on its lit faces, with 15% darker tonal shift on shadowed faces. A subtle 8%-opacity outer glow halo surrounds the subject. Small supporting context: three thin connecting lines in accent vivid cyan `#06B6D4` arcing from the subject toward the canvas edges (suggesting connectivity), and a soft 8% drop shadow grounding the subject. Background is deep secondary navy `#0A0E27` (about 30% of canvas, including shadowed plane). The subject is clearly the singular focal element. Composed as a 600×600 hero block with 15% padding around the subject. NO text, letters, numbers, or labels anywhere. Color values are rendering guidance only.++> **B — corporate-photo + cool-corporate executive headshot, text_policy: none, 600×800**+>+> Editorial corporate portrait photograph of one professional executive. The person is centered slightly left of canvas center, photographed from chest-up at eye level, looking confidently toward the camera with a relaxed natural expression — not posed-stiff, not over-smiling. Professionally attired in a contemporary business setting (a tailored blazer, neutral palette clothing). Soft natural light from the upper left, gentle shadow on the right side of the face. Diverse, professionally attired subject, photorealistically rendered, contemporary styling. Background is a softly out-of-focus office context — secondary light gray `#F8F9FA` wall with a subtle hint of primary deep navy `#1E3A5F` in a blurred architectural element. Color grading is cool-corporate — restrained, professional. Shallow depth of field — subject sharp, background gently blurred. Subject's eyes positioned at the upper-third horizontal line. Composed as a 600×800 bio portrait with 10% padding. NO text, name tags, or captions in the image. Color values are rendering guidance only.++> **C — ink-notes + mono-ink big-number stat, text_policy: embedded, 800×500**+>+> Professional hand-drawn visual-note style on pure white background. The image's central content is the hand-lettered number "100x" — rendered in bold confident ink strokes occupying about 50% of the canvas height, centered with deliberate slight wobble characteristic of hand-lettering. Text is in English/Latin characters only. Beneath the number, a thin hand-drawn underline in ink. To the side of the number, one small hand-drawn doodle decoration — a star or upward arrow — adds visual rhythm. Accent coral `#E8655A` (from the deck's accent) appears only as a tiny emphasis dot, totaling under 4% of canvas. Background is pure white `#FFFFFF`. Composed as an 800×500 typographic hero block with 20% padding around the number. No other text or labels in the image — just the "100x" headline and the small doodle.++> **D — vector-illustration + cool-corporate cover background, text_policy: none, 1280×720**+>+> Clean flat vector illustration backdrop. Atmospheric composition with no central subject — bold geometric shapes arranged along the canvas edges to leave the center calm. Primary deep navy `#1E3A5F` forms a confident diagonal block across the lower-left third; secondary light gray `#F8F9FA` occupies the upper two-thirds as breathing space; accent gold `#D4AF37` appears only as one thin geometric line near the lower right corner (under 5% of canvas). Crisp 2px outlines, no gradients, single 8% soft drop shadow under the navy block. The central 60% of the canvas is deliberately calm and unbusy — designed to receive a slide title overlaid in SVG. Composed as a 1280×720 full-bleed PPT background. NO text, letters, numbers, signs, watermarks, or written symbols anywhere in the image. Color values are rendering guidance only — do not display HEX codes or color names as text. Simplified geometric shapes only.+---## 5. Global Hard Rules@@ -228,7 +274,20 @@ The table below is **a reference for the one case where you want the in-image le**When to use the table**: a designed title (cover main title, chapter heading) on a deck whose visual identity is grounded in the SVG body typography, and where a surprise font choice would feel out of place.-**CJK note**: most image models render Chinese characters poorly. For embedded text on a CJK deck, prefer English in the image or accept malformed glyphs.+**In-image text accuracy — what image models can and cannot render reliably**++Image models have **variable accuracy on text rendering**. Pick the in-image text content with this in mind, especially for §4.1 Primitive C (typographic hero) where the word *is* the image:++| Content | Reliability |+|---|---|+| English short text (1-5 characters, one short word) | High — most modern models render correctly most of the time |+| English longer text (full sentences) | Low — typos, glyph errors, dropped letters |+| Simple numbers / symbols (`100`, `5x`, `$50M`, `∞`, `?`) | Variable — verify after generation, often fail at large display scale |+| Chinese / Japanese / Korean characters | Very low — most models fail consistently |++**Prefer in-image**: a short English word (1-2 words max), a simple number, a single symbol or letter.++**Push to SVG overlay instead**: long quotes / sentences, CJK headlines, complex composite text. When the desired headline is long or CJK, switch to **Primitive D (atmospheric backdrop)** and overlay the headline as SVG text.### 5.4 No brand names or trademarks in the subject@@ -255,14 +314,24 @@ Write `project/images/image_prompts.json` with this shape:{"filename": "cover_bg.png","purpose": "Cover background (Slide 01)",- "type": "background",- "page_role": "local",+ "page_role": "hero_page","text_policy": "none","aspect_ratio": "16:9","image_size": "2K",- "prompt": "{fully assembled paragraph per §4}",+ "prompt": "{fully assembled paragraph per §4 — use §4.1 Primitive D for atmospheric cover}","alt_text": "Modern tech abstract background with deep blue gradient and digital waves","status": "Pending"+ },+ {+ "filename": "framework_p05.png",+ "purpose": "Methodology framework (Slide 05)",+ "type": "framework",+ "page_role": "local",+ "text_policy": "none",+ "aspect_ratio": "4:3",+ "image_size": "1K",+ "prompt": "{fully assembled paragraph per §4}",+ "status": "Pending"}]}@@ -276,7 +345,7 @@ Write `project/images/image_prompts.json` with this shape:| `deck_palette` | yes | Step 2 lock | Single palette name shared by all items || `color_scheme` | yes | `design_spec.md §III` | HEX triplet from Strategist || `items[].filename` | yes | `§VIII` resource list | Output filename with extension |-| `items[].type` | yes | Step 3 per-image | One of: `background`, `hero`, `portrait`, `typography`, `infographic`, `flowchart`, `framework`, `matrix`, `cycle`, `funnel`, `pyramid`, `comparison`, `timeline`, `map`, `scene` |+| `items[].type` | conditional | Step 3 per-image (only when `page_role: local`) | One of 11 internal-composition types: `infographic`, `flowchart`, `framework`, `matrix`, `cycle`, `funnel`, `pyramid`, `comparison`, `timeline`, `map`, `scene`. **Omit `type` entirely when `page_role: hero_page`** — the composition comes from §4.1 primitives written directly into the prompt, not from a type file. || `items[].page_role` | yes | Step 3 per-image | `local` (default — region block on SVG page) or `hero_page` (image is page's main voice; SVG overlay minimal or empty) || `items[].text_policy` | yes | Step 3 per-image | `none` (default — no text in image) or `embedded` (image contains decorative lettering, designed title, or hand-lettered keywords) || `items[].aspect_ratio` | yes | Container sizing | Passed to `image_gen.py --aspect_ratio` |@@ -285,6 +354,8 @@ Write `project/images/image_prompts.json` with this shape:| `items[].alt_text` | no | Accessibility | Short caption || `items[].status` | yes | CLI manages | `Pending` initially; CLI updates to `Generated` / `Failed` / `Needs-Manual` |+> **Back-compat for legacy `type` values**: existing manifests using `background` / `hero` / `portrait` / `typography` (the four removed pseudo-types) remain readable. Read them as: `background` → `page_role: hero_page` + no type; `hero` → `page_role: hero_page` + no type (use §4.1 Primitive A in prompt); `portrait` → `page_role: local` + no type (use §4.1 Primitive B); `typography` → `page_role: hero_page` + `text_policy: embedded` + no type (use §4.1 Primitive C). New manifests should follow the rule above (omit `type` when `page_role: hero_page`).+>> Existing manifests without `deck_rendering` / `deck_palette` / `type` / `page_role` / `text_policy` remain valid — older items default to `page_role: local`, `text_policy: none`. Legacy `page_role: full_page` (pre-2026-05-15) is read as `hero_page`.---@@ -426,14 +497,15 @@ When the Resource List row has no `Reference`, infer a reasonable image from `Pu| Purpose | A reasonable starting point ||---------|-----------------------------|-| Cover | `type: background` or `hero`; choose page_role and text_policy by what the cover should communicate |-| Chapter divider | `type: background` or `hero`; often `hero_page` with a designed chapter title |-| Methodology / framework illustration | `type: framework` |-| Process / workflow illustration | `type: flowchart` |-| Before/After or two-option page | `type: comparison` |-| Team / lifestyle photo | `type: scene`; rendering = `corporate-photo` or `warm-scene` |-| Big-number / hero quote block | `type: typography` or `hero`; often `hero_page` |-| Mood transition / atmosphere | `type: scene` or `background`; often `hero_page` |+| Cover | `page_role: hero_page` + §4.1 Primitive A (single-subject) or D (atmospheric); choose `text_policy` by what the cover should communicate |+| Chapter divider | `page_role: hero_page` + Primitive D (atmospheric) or A (single-subject); often `text_policy: embedded` with a designed chapter title |+| Methodology / framework illustration | `type: framework`, `page_role: local` |+| Process / workflow illustration | `type: flowchart`, `page_role: local` |+| Before/After or two-option page | `type: comparison`, `page_role: local` |+| Team / lifestyle photo (group) | `type: scene`, `page_role: local`; rendering = `corporate-photo` or `warm-scene` |+| Single-person headshot / bio | `page_role: local` + §4.1 Primitive B (portrait); rendering = `corporate-photo` for photo realism |+| Big-number / hero quote block | `page_role: hero_page` + §4.1 Primitive C (typographic); `text_policy: embedded` |+| Mood transition / atmosphere | `page_role: hero_page` + Primitive D (atmospheric), or `type: scene` if narrative |### When Images Are Unsatisfactorydiff --git a/skills/ppt-master/references/image-type-templates/_index.md b/skills/ppt-master/references/image-type-templates/_index.mdindex d3aed5d4..4347fcaa 100644--- a/skills/ppt-master/references/image-type-templates/_index.md+++ b/skills/ppt-master/references/image-type-templates/_index.md@@ -1,32 +1,36 @@# Type Templates — Index-A **type** describes what the image's **internal composition** looks like — its layout skeleton. Type is decided **per image**, not per deck (one deck typically uses 2-5 different types).+A **type** describes the **internal geometric composition skeleton** of a local infographic image block — what the layout looks like *inside the rectangle the model paints*. Type is decided **per image**, not per deck (one deck typically uses 2-4 different types).-> **Type vs PPT page layout — important**: type describes the composition *inside* the AI image block. The PPT slide's overall layout (where the image sits, how SVG text wraps around it) is a separate decision made by Strategist and Executor. Type only governs what's inside the rectangle the model paints.->-> **Type is orthogonal to `page_role`**: every type's container guidance assumes `page_role: local` (region block on an SVG page). When promoted to `page_role: hero_page`, the same type's composition fills the slide with SVG kept minimal above. See [`image-generator.md`](../image-generator.md) §1.+## What Type *is* and *is not*++**Type *is*** — the geometric skeleton of a local infographic block. Each type has a non-interchangeable layout structure: a 2×2 grid is not a closed loop; an upward-narrowing pyramid is not a top-wide funnel; a linear timeline is not a sequential flowchart. The skeleton constrains where elements go.++**Type *is not***:++- *not* "what this image is for in the PPT page" — that's `page_role` (`local` vs `hero_page`, see [`image-generator.md`](../image-generator.md) §1)+- *not* "what subject occupies the image" — single subject, single person, big number, no subject: these are all expressible through §4.1 hero-page primitives or natural-language prompt description, not through types+- *not* "what genre the image is" — Background / Photography / Illustration / Diagram / Decorative is a narrative shorthand in `design_spec.md §VIII Genre`, not the same vocabulary as `image_type`++**When to skip type entirely** — when `page_role: hero_page` (the image is the page's main voice: cover, chapter divider, mood transition, signature stat, closing quote), do **not** pick a type. Instead describe the composition directly using the four primitives in [`image-generator.md`](../image-generator.md) §4.1 (single-subject / portrait / typographic / atmospheric). The 11 types below are for local infographic blocks only.----## 1. Catalog (15 types)+## 1. Catalog (11 types)Each type has its own file with: composition skeleton (LAYOUT / ELEMENTS / NEGATIVE SPACE), text-policy variants, and a fewshot prompt snippet.| Type | Internal composition | Typical use ||---|---|---|-| [`background`](./background.md) | Atmospheric backdrop, no central subject | Cover bg, chapter divider bg, hero overlay bg |-| [`hero`](./hero.md) | Single dominant subject (60-70% of canvas) | Product/concept reveal, chapter title bg |-| [`portrait`](./portrait.md) | One person headshot / upper body, neutral background | Team page / speaker bio / testimonial / founder profile |-| [`typography`](./typography.md) | Large headline / number / single word as visual | Big-stat pages, slogan pages, chapter openers |-| [`infographic`](./infographic.md) | 2-5 ordered zones with icons + minimal labels | Data summary / step list / KPI rundown |-| [`flowchart`](./flowchart.md) | Sequential blocks connected by arrows | Process / workflow / pipeline |-| [`framework`](./framework.md) | Central node + radiating satellites | Methodology / model / system architecture |+| [`infographic`](./infographic.md) | 2-5 parallel ordered zones with icons + minimal labels | Data summary / step list / KPI rundown |+| [`flowchart`](./flowchart.md) | Sequential blocks connected by directional arrows | Process / workflow / pipeline |+| [`framework`](./framework.md) | Central node + radiating satellites (hub-spoke) | Methodology / model / system architecture || [`matrix`](./matrix.md) | 2×2 quadrant grid with two perpendicular axes | SWOT / BCG / Eisenhower / Ansoff / Porter |-| [`cycle`](./cycle.md) | Closed loop, 3-6 steps with arrows returning | PDCA / flywheel / design thinking / continuous improvement |+| [`cycle`](./cycle.md) | Closed loop, 3-6 steps with arrows returning to start | PDCA / flywheel / design thinking / continuous improvement || [`funnel`](./funnel.md) | Top-wide bottom-narrow conversion stack | Marketing funnel / sales pipeline / hiring funnel || [`pyramid`](./pyramid.md) | Bottom-wide top-narrow hierarchical tiers | Maslow / capability stack / value hierarchy |-| [`comparison`](./comparison.md) | Split composition (left vs right, before vs after) | A/B / pros-cons / Before-After |-| [`timeline`](./timeline.md) | Linear progression along an axis | History / roadmap / evolution |+| [`comparison`](./comparison.md) | Symmetric split (left vs right, before vs after) | A/B / pros-cons / Before-After |+| [`timeline`](./timeline.md) | Linear axis with milestone markers | History / roadmap / evolution || [`map`](./map.md) | Stylized geographic outline with annotated markers | Offices / market presence / regional data / supply chain || [`scene`](./scene.md) | Atmospheric environment with narrative | Story / lifestyle / case study |@@ -34,14 +38,10 @@ Each type has its own file with: composition skeleton (LAYOUT / ELEMENTS / NEGAT## 2. Auto-selection — per-image `Purpose` → type-For each row in `design_spec.md §VIII Image Resource List`, match `Purpose` against this table.+For each row in `design_spec.md §VIII Image Resource List` where `page_role: local`, match `Purpose` against this table.| `Purpose` keyword | Type ||---|---|-| Cover background / chapter background / divider bg | `background` |-| Product reveal / hero / headline visual | `hero` |-| Team headshot / speaker bio / founder profile / testimonial | `portrait` |-| Big number / KPI single visual / quote | `typography` || Data summary / metrics rundown / step list | `infographic` || Process / workflow / pipeline / steps with arrows | `flowchart` || Methodology / model / framework / architecture diagram | `framework` |@@ -52,7 +52,8 @@ For each row in `design_spec.md §VIII Image Resource List`, match `Purpose` aga| Comparison / Before-After / A/B / VS | `comparison` || History / evolution / roadmap / timeline | `timeline` || Offices / market presence / regions / supply chain / geography | `map` |-| Team / lifestyle / story / scenario / case | `scene` |+| Team / lifestyle / story / scenario / case (group, with environment) | `scene` |+| Cover / chapter divider / mood transition / big number / hero quote / single-subject hero / single-person headshot | **No type — use `page_role: hero_page` + [`image-generator.md`](../image-generator.md) §4.1 primitives** |`text_policy` and `page_role` are decided per image — see each type file's variants section and the page's communication goal.@@ -64,10 +65,6 @@ The Resource List's `Dimensions` column is authoritative. If absent, use these d| Type | Default container | Aspect ratio ||---|---|---|-| background | 1280×720 full-bleed; `page_role: local` when SVG text overlays on top, `page_role: hero_page` when image alone is the page | 16:9 |-| hero | 800×600 or 1280×720 | 4:3 / 16:9 |-| portrait | 400×500 or 600×800 | 4:5 / 3:4 |-| typography | 800×500 | 16:10 || infographic | 600×500 or 700×700 | ~1.2 / 1 || flowchart | 1200×400 (horizontal banner) | 3:1 || framework | 700×700 square | 1:1 |@@ -80,12 +77,15 @@ The Resource List's `Dimensions` column is authoritative. If absent, use these d| map | 1280×720 or 1200×500 | 16:9 / 2.4:1 || scene | 1200×720 wide / 800×600 | 16:10 / 4:3 |+For `page_role: hero_page` images, default container is the slide canvas (e.g. 1280×720 for 16:9) — see §4.1 primitives in [`image-generator.md`](../image-generator.md).+---## 4. How to use-1. Pick the type per row in the Image Resource List using the table above.-2. `read_file image-type-templates/<type>.md` — only the types actually used in this deck. Most decks use 2-4 types; load each at most once.-3. Apply the type's composition skeleton when assembling the prompt, alongside the locked deck-wide rendering and palette.+1. For each `page_role: local` row in the Image Resource List, pick the type using the auto-selection table above.+2. For each `page_role: hero_page` row, **skip type selection** — go straight to [`image-generator.md`](../image-generator.md) §4.1 primitives.+3. `read_file image-type-templates/<type>.md` — only the types actually used in this deck. Most decks use 2-4 types; load each at most once.+4. Apply the type's composition skeleton when assembling the prompt, alongside the locked deck-wide rendering and palette.**Multiple types per deck is normal.** Locking is at rendering + palette level, not type level.diff --git a/skills/ppt-master/references/image-type-templates/background.md b/skills/ppt-master/references/image-type-templates/background.mddeleted file mode 100644index 5ebae082..00000000--- a/skills/ppt-master/references/image-type-templates/background.md+++ /dev/null@@ -1,50 +0,0 @@-# Type: background--Atmospheric backdrop with **no central subject**. The image's job is to set tone and provide a calm field for SVG text overlay. The most common type for PPT covers, chapter dividers, and breathing pages.--> **What background means inside a PPT block**: the image is **all atmosphere, no subject**. Unlike `hero`, there is no dominant figure. Unlike `infographic`, there are no labeled zones. Background is pure ambience.--## 1. Composition skeleton--```- ┌────────────────────────────────────┐- │ ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ │- │ ░ atmospheric gradients ░ │- │ ░ + subtle geometric anchor ░ │- │ ░ (corner / edge, not center) ░ │- │ ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ │- └────────────────────────────────────┘- ↑- Center 60% is intentionally calm- (SVG title will overlay here)-```--| LAYOUT | Atmospheric field — gradient, subtle pattern, or restrained color block composition. No dominant subject |-| ELEMENTS | Optional small geometric anchor in a corner or along an edge (a line, a small shape, a subtle motif). Never centered |-| NEGATIVE SPACE | The center 60-70% must be calm — low visual weight, ready to receive SVG title/text overlay |-| TEMPERAMENT | Sets the deck's overall tone: corporate-restrained, warm-friendly, tech-energetic, etc. |--## 2. Text-policy variants--### `text_policy: none`--Backgrounds are atmospheric — for most uses, keep text out and let SVG handle overlay copy.--Sample fragment:--> NO text of any kind anywhere in the image — no letters, numbers, signs, watermarks, labels, or written symbols. The image is pure atmosphere; SVG text overlay will be added externally.--### `text_policy: embedded` (decorative or designed lettering)--When the background's mood is enhanced by lettering — large decorative word in the bleed, retro stamp, scattered alphabet texture, or a designed cover title that's part of the artwork. Name the text role explicitly in the prompt:--- Decorative lettering: "large decorative 'GROWTH' lettering in 3D extruded retro style as background element; spelling not critical"-- Designed cover title: "main title text '{exact words}' rendered in {font family echoing deck's body typography}; content must be accurate"-------## 3. Fewshot prompt snippets--**Snippet A — vector-illustration + cool-corporate cover background, text_policy: none, 1280×720**--> Clean flat vector illustration backdrop. Atmospheric composition with no central subject — bold geometric shapes arranged along the canvas edges to leave the center calm. Primary deep navy `#1E3A5F` forms a confident diagonal block across the lower-left third; secondary light gray `#F8F9FA` occupies the upper two-thirds as breathing space; accent gold `#D4AF37` appears only as one thin geometric line near the lower right corner (under 5% of canvas). Crisp 2px outlines, no gradients, single 8% soft drop shadow under the navy block. The central 60% of the canvas is deliberately calm and unbusy — designed to receive a slide title overlaid in SVG. Composed as a 1280×720 full-bleed PPT background. NO text, letters, numbers, signs, watermarks, or written symbols anywhere in the image. Color values are rendering guidance only — do not display HEX codes or color names as text. Simplified geometric shapes only.\ No newline at end of filediff --git a/skills/ppt-master/references/image-type-templates/hero.md b/skills/ppt-master/references/image-type-templates/hero.mddeleted file mode 100644index 15f75b2c..00000000--- a/skills/ppt-master/references/image-type-templates/hero.md+++ /dev/null@@ -1,44 +0,0 @@-# Type: hero--Single dominant subject occupying 60-70% of the canvas. The hero image *anchors* a slide — it's what the page is *about* visually. Used for product reveals, concept introductions, chapter title visuals, brand statement pages.--> **What hero means inside a PPT block**: one confident dominant subject with intentional supporting context. Unlike `background` (no subject), hero is **all about the subject**. Unlike `infographic` (multiple labeled zones), hero is **singular focal subject**.--## 1. Composition skeleton--```- ┌────────────────────────────────────┐- │ │- │ ┌──────────────┐ │- │ │ │ │- │ │ HERO │ │- │ │ SUBJECT │ (60-70% │- │ │ │ of canvas) │- │ └──────────────┘ │- │ │- │ small supporting context only │- └────────────────────────────────────┘-```--| LAYOUT | One dominant subject occupying 60-70% of the canvas area, positioned with intent (centered, slight left/right offset, or rule-of-thirds) |-| ELEMENTS | The hero subject + minimal supporting context (background environment, small accent elements). Supporting elements <30% of canvas |-| NEGATIVE SPACE | Generous around the subject — at least 15% padding on the subject's "open" side |-| BALANCE | Subject's visual weight clearly dominant; no second-place subject competing |--## 2. Text-policy variants--### `text_policy: none` (most common)--The hero subject is the visual; any title or label comes from SVG overlay.--### `text_policy: embedded`--The hero subject itself includes text — product name on packaging, a hand-lettered word as part of the subject, a designed title floating beside the figure. Specify font family in the prompt to echo the deck's body typography. If the headline *is* the entire visual, switch to `typography` type.-------## 3. Fewshot prompt snippets--**Snippet A — 3d-isometric + tech-neon product reveal, text_policy: none, 600×600**--> 3D isometric illustration in true 30°/30°/30° projection. One dominant product-form subject — a stylized device or sleek tech object — occupies the center of the canvas at roughly 65% of the area. The subject is rendered in primary electric blue `#0EA5E9` on its lit faces, with 15% darker tonal shift on shadowed faces. A subtle 8%-opacity outer glow halo surrounds the subject. Small supporting context: three thin connecting lines in accent vivid cyan `#06B6D4` arcing from the subject toward the canvas edges (suggesting connectivity), and a soft 8% drop shadow grounding the subject. Background is deep secondary navy `#0A0E27` (about 30% of canvas, including shadowed plane). The subject is clearly the singular focal element. Composed as a 600×600 half-page hero block with 15% padding around the subject. NO text, letters, numbers, or labels anywhere. Color values are rendering guidance only.\ No newline at end of filediff --git a/skills/ppt-master/references/image-type-templates/portrait.md b/skills/ppt-master/references/image-type-templates/portrait.mddeleted file mode 100644index 10cbaf81..00000000--- a/skills/ppt-master/references/image-type-templates/portrait.md+++ /dev/null@@ -1,52 +0,0 @@-# Type: portrait--A **single person headshot or upper-body shot** — frontal or three-quarter turn, neutral or minimal background, the face as the visual anchor. Used for team pages, speaker bios, founder profiles, testimonial pages, "about us" sections, executive introductions.--> **What portrait means inside a PPT block**: **one person, face-forward, isolated from environment**. Unlike `scene` (atmospheric environment with narrative), portrait strips away context to focus on the person. Unlike `hero` (any singular subject — product / object / person), portrait is specifically **human face / upper body**.-------## 1. Composition skeleton--```- ┌────────────────────────────────────┐- │ │- │ ┌─────────┐ │- │ │ Head │ │- │ │ / Face │ │- │ ├─────────┤ │- │ │ Shoulders│ │- │ │ / Torso │ │- │ └─────────┘ │- │ │- │ Subject occupies 50-65% of │- │ canvas height, centered │- └────────────────────────────────────┘-```--| LAYOUT | Single person centered (or rule-of-thirds offset); head occupies upper third, torso/shoulders the middle. Background is neutral, minimal, or softly blurred. |-| ELEMENTS | The person (face + upper body) + minimal background context (a soft color field, a gentle gradient, an out-of-focus environment cue). No competing foreground objects. |-| NEGATIVE SPACE | Generous around the head — at least 15% padding above the crown of the head |-| BALANCE | The face is the visual anchor; eyes positioned at approximately the upper-third horizontal line (classical portrait rule) |--## 2. Text-policy variants--### 3.1 `text_policy: none`--The portrait shows the person only; names, titles, and quotes are handled in SVG.--Sample fragment:--> NO text, letters, numbers, name tags, or captions in the image. The portrait is the person only.--### 3.2 `text_policy: embedded`--If text is part of the page's visual (a name carved as a design element, a quote alongside the figure), the type is likely `hero` or `typography` rather than `portrait`.-------## 3. Fewshot prompt snippets--**Snippet A — corporate-photo + cool-corporate executive headshot, text_policy: none, 600×800**--> Editorial corporate portrait photograph of one professional executive. The person is centered slightly left of canvas center, photographed from chest-up at eye level, looking confidently toward the camera with a relaxed natural expression — not posed-stiff, not over-smiling. Professionally attired in a contemporary business setting (a tailored blazer, neutral palette clothing). Soft natural light from the upper left, gentle shadow on the right side of the face. Diverse, professionally attired subject, photorealistically rendered, contemporary styling. Background is a softly out-of-focus office context — secondary light gray `#F8F9FA` wall with a subtle hint of primary deep navy `#1E3A5F` in a blurred architectural element. Color grading is cool-corporate — restrained, professional. Shallow depth of field — subject sharp, background gently blurred. Subject's eyes positioned at the upper-third horizontal line. Composed as a 600×800 half-page bio portrait with 10% padding. NO text, name tags, or captions in the image. Color values are rendering guidance only.\ No newline at end of filediff --git a/skills/ppt-master/references/image-type-templates/typography.md b/skills/ppt-master/references/image-type-templates/typography.mddeleted file mode 100644index 925cea72..00000000--- a/skills/ppt-master/references/image-type-templates/typography.md+++ /dev/null@@ -1,73 +0,0 @@-# Type: typography--A large headline, number, or single word as the primary visual. The text itself is the image. Used for big-stat pages ("$2.3M"), slogan pages ("MOVE FAST"), chapter openers, hero quotes, signature numbers in consulting decks.--> **What typography means inside a PPT block**: the image's primary content is a piece of text rendered as art. Unlike all other types where text is forbidden or minimal, typography **requires** text — and that text is the visual anchor.--## 1. Composition skeleton--```- ┌────────────────────────────────────┐- │ │- │ ┌─────────────┐ │- │ │ │ │- │ │ $2.3M │ ← The │- │ │ │ text │- │ └─────────────┘ IS the │- │ image │- │ small supporting visual │- │ │- └────────────────────────────────────┘-```--| LAYOUT | Large text element occupying 40-60% of canvas height, centered or rule-of-thirds positioned |-| ELEMENTS | The text + minimal supporting visual context (small icon, geometric anchor, accent line). Supporting elements <25% of weight |-| NEGATIVE SPACE | Generous around the text — at least 20% padding |-| TEXT CHARACTER | Rendered with intention — confident weight, deliberate spacing, may have texture/style matching deck rendering |-------## 2. Critical: text accuracy--Image models have **variable accuracy on text rendering**:--- **English short text (1-5 characters / one short word)**: most modern models render correctly most of the time-- **English longer text (sentences)**: high failure rate (typos, glyph errors)-- **Numbers / symbols**: variable, often fail at typography scale-- **Chinese / Japanese / Korean characters**: most models fail consistently--**For typography type, prefer**:-- A short English word (1-2 words max)-- A simple number ("100", "5x", "$50M") — verify after generation-- A simple symbol or letter ("∞", "?", "A")--**Avoid for typography type**:-- Long quotes / sentences (use SVG text instead)-- Chinese/CJK characters (model failure expected)-- Complex symbols or composite text--When the desired headline is long or CJK: switch to `background` type and overlay the headline as SVG text.--## 3. Text-policy variants--### `text_policy: embedded`--typography type is built around in-image text — the headline word, number, or quote is the visual. `text_policy: embedded` is the natural choice.--Sample fragment:--> The image's central content is the hand-lettered (or appropriately styled) word "GROWTH" — rendered as a large confident headline occupying about 50% of the canvas height, in {font family echoing deck's body typography}. Text is in English only (most models render CJK characters poorly). No other text or labels in the image — just the single headline word.--**Font family in prompt**: name the family so the lettering coheres with the deck's SVG typography. Read `spec_lock.md deck_typography` if present; otherwise infer from rendering (vector-illustration / flat → clean sans-serif; editorial → serif; sketch-notes / ink-notes → hand-written; 3d-isometric / digital-dashboard → geometric display).--### `text_policy: none`--If the page wants no text inside the image, the type is probably `background` or `hero`, not `typography`.-------## 4. Fewshot prompt snippets--**Snippet A — ink-notes + mono-ink big-number stat, text_policy: embedded, 800×500**--> Professional hand-drawn visual-note style on pure white background. The image's central content is the hand-lettered number "100x" — rendered in bold confident ink strokes occupying about 50% of the canvas height, centered with deliberate slight wobble characteristic of hand-lettering. Text is in English/Latin characters only. Beneath the number, a thin hand-drawn underline in ink. To the side of the number, one small hand-drawn doodle decoration — a star or upward arrow — adds visual rhythm. Accent coral `#E8655A` (from the deck's accent) appears only as a tiny emphasis dot above the number's exclamation, totaling under 4% of canvas. Background is pure white `#FFFFFF`. Composed as an 800×500 typography block with 20% padding around the number. No other text or labels in the image — just the "100x" headline and the small doodle.\ No newline at end of filediff --git a/skills/ppt-master/references/strategist.md b/skills/ppt-master/references/strategist.mdindex 026aa164..4aa684ef 100644--- a/skills/ppt-master/references/strategist.md+++ b/skills/ppt-master/references/strategist.md@@ -401,7 +401,7 @@ After the user picks a candidate, scan the outline and surface any pages where t| Layout suggestion | e.g., `Wide landscape (suitable for full-screen/illustration)` || **Layout pattern** | **MANDATORY** — one or more `#<id> <name>` joined by ` + ` from `image-layout-patterns.md`. Combine a Primary id with optional Modifier ids when the page needs it (e.g. `#48 side-by-side comparison + #21 rounded rectangle crop + #29 two-stop scrim`). A single Primary is fine when the page calls for it. See the GATE earlier in this section. Empty cells or invented ids are invalid. || Purpose | e.g., `Cover background` |-| Type | Narrative shorthand: Background / Photography / Illustration / Diagram / Decorative pattern. (The internal-composition type used by Image_Generator — one of `background / hero / typography / infographic / flowchart / framework / comparison / timeline / scene` — is inferred from `Purpose` per [`image-type-templates/_index.md`](./image-type-templates/_index.md); no need to label every row.) |+| Genre | Narrative shorthand: Background / Photography / Illustration / Diagram / Decorative pattern. (The internal-composition `image_type` used by Image_Generator — one of 11 types: `infographic / flowchart / framework / matrix / cycle / funnel / pyramid / comparison / timeline / map / scene` — is inferred from `Purpose` per [`image-type-templates/_index.md`](./image-type-templates/_index.md), only for `page_role: local` rows. For `page_role: hero_page` rows, no type is selected; Image_Generator uses §4.1 primitives directly.) || **Acquire Via** | `ai` / `web` / `user` / `placeholder` — drives Step 5 dispatch || Status | Initial status must be `Pending`, `Existing`, or `Placeholder`; see [`svg-image-embedding.md`](svg-image-embedding.md) for the full status enum || **Reference** | Free-form **intent description** (NOT a search query); feeds Image_Generator (ai) or Image_Searcher (web) |@@ -442,10 +442,10 @@ After the user picks a candidate, scan the outline and surface any pages where t| `ai` | "Atmospheric backdrop suggesting digital innovation; calm central area reserved for slide title overlay; light geometric anchor at one edge" || `ai` | "Four-stage value chain from raw input to R&D output; icons should suggest tax-form → cost-reduction → equipment-upgrade → innovation; no text labels (SVG overlays them)" |-**Image type descriptions**:+**Image Genre descriptions** (narrative shorthand for the §VIII `Genre` column):-| Type | Suitable Scenarios |-|------|-------------------|+| Genre | Suitable Scenarios |+|-------|-------------------|| Background | Full-page backgrounds for covers/chapter pages; reserve text area || Photography | Real scenes, people, products, architecture || Illustration | Flat design, vector style, concept diagrams |diff --git a/skills/ppt-master/templates/design_spec_reference.md b/skills/ppt-master/templates/design_spec_reference.mdindex ff122000..012f0eca 100644--- a/skills/ppt-master/templates/design_spec_reference.md+++ b/skills/ppt-master/templates/design_spec_reference.md@@ -240,8 +240,8 @@ Catalog read: 71 templates## VIII. Image Resource List (if needed)-| Filename | Dimensions | Ratio | Purpose | Type | Layout pattern | Acquire Via | Status | Reference | text_policy | page_role |-| -------- | --------- | ----- | ------- | ---- | -------------- | ----------- | ------ | --------- | ----------- | --------- |+| Filename | Dimensions | Ratio | Purpose | Genre | Layout pattern | Acquire Via | Status | Reference | text_policy | page_role |+| -------- | --------- | ----- | ------- | ----- | -------------- | ----------- | ------ | --------- | ----------- | --------- || cover_bg.png | {canvas_info['dimensions']} | [ratio] | Cover background | Background | #1 full-bleed background with floating title + #29 two-stop scrim | ai | Pending | [subject + intent + composition, no style/HEX] | | |> **Layout pattern column is MANDATORY** — value is one or more `#<id> <name>` joined by ` + ` drawn verbatim from [`references/image-layout-patterns.md`](../references/image-layout-patterns.md) (Primary + optional Modifiers). Empty cells, paraphrased names, or invented ids invalidate the row. See `strategist.md §h` GATE for the three-layer requirement (read → produce → image-as-canvas coverage).@@ -252,7 +252,7 @@ Catalog read: 71 templates- **Existing** — user-supplied, place in `images/`- **Placeholder** — not yet processed, use dashed border in SVG-**Type** (narrative shorthand — kept for backward compatibility; Image_Generator infers its 9-way internal-composition type from `Purpose`):+**Genre** (narrative shorthand for the row — describes what kind of asset this is, not the geometric composition. Image_Generator infers its 11-way internal-composition `image_type` from `Purpose` independently; for `page_role: hero_page` rows, no type is needed at all — see [`references/image-generator.md`](../references/image-generator.md) §4.1 primitives):- **Background** — full-page (covers / chapters); reserve text area- **Photography** — real scenes, people, products, architecture