bggg-creator-image2ppt

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

BGGG Creator Image2PPT

用这个 skill 把平面视觉稿转换成可编辑的

.pptx

。核心设计是“先形成结构化 manifest，再由脚本生成 PPTX”。

二进制图片输入：Codex 负责识别版式、文字和组件；默认调用
```
imagegen
```
skill 生成或清理背景、装饰、产品、图表等组件图片；文本尽量还原为 PowerPoint 文本框。
HTML/SVG 输入：优先把代码结构解析为原生 PPT 元素；复杂节点再降级为图片组件。
所有中间图片、manifest、诊断和输出都必须放在本 skill 的
```
projects/YYYYMMDD_slug/
```
目录里。

Use this skill to convert flat visual drafts into editable

.pptx

files. The core design is "first form a structured manifest, then generate PPTX via scripts."

Binary image input: Codex is responsible for identifying layout, text and components; it defaults to calling the
```
imagegen
```
skill to generate or clean component images such as backgrounds, decorations, products, charts, etc.; text is restored to PowerPoint text boxes as much as possible.
HTML/SVG input: Prioritize parsing code structures into native PPT elements; complex nodes are downgraded to image components.
All intermediate images, manifests, diagnostics and outputs must be placed in the
```
projects/YYYYMMDD_slug/
```
directory of this skill.

Codex 默认策略

Codex Default Strategy

在 Codex 中处理 PNG/JPEG/WebP 等二进制图片时，默认执行这个顺序：

使用 Codex 视觉能力观察源图，列出幻灯片尺寸、背景、标题、正文、图标、照片、图表、装饰、阴影、遮罩等对象。
对于可编辑文字，直接写入 manifest 的
```
text
```
元素，不要把文字只做成图片。
对于背景、照片、图标、复杂插画、图表、纹理、遮挡后需要补全的背景，使用
```
imagegen
```
skill 生成、清理或重建独立组件图片。
把
```
imagegen
```
生成的图片复制到当前项目目录的
```
imagegen_assets/
```
或
```
component_images/
```
，不要让项目依赖
```
$CODEX_HOME
```
的临时输出。

写

manifest.json

，用

scripts/image2pptx.py build

生成 PPTX。

验证 PPTX 能被
```
python-pptx
```
重新打开，并记录图层/文本/图片数量和已知限制。

如果用户明确要求“不重绘”“保持原图像素”，可以用原图裁切或全画布透明 PNG 作为组件图片；但默认仍要用 Codex imagegen 能力辅助背景清理、缺失区域补全和组件干净化。

When processing binary images such as PNG/JPEG/WebP in Codex, the following sequence is executed by default:

Use Codex's visual capability to observe the source image, list objects such as slide size, background, title, body text, icons, photos, charts, decorations, shadows, masks, etc.
For editable text, directly write it into the
```
text
```
element of the manifest; do not only render text as images.
For backgrounds, photos, icons, complex illustrations, charts, textures, and backgrounds that need to be completed after occlusion, use the
```
imagegen
```
skill to generate, clean or rebuild independent component images.
Copy the images generated by
```
imagegen
```
to the
```
imagegen_assets/
```
or
```
component_images/
```
directory of the current project; do not let the project depend on temporary outputs from
```
$CODEX_HOME
```
.

Write

manifest.json

, then use

scripts/image2pptx.py build

to generate PPTX.

Verify that the PPTX can be reopened by
```
python-pptx
```
, and record the number of layers/text/images and known limitations.

If the user explicitly requests "no redrawing" or "retain original pixels", the original image can be cropped or used as a full-canvas transparent PNG as a component image; but by default, Codex imagegen capability should still be used to assist background cleaning, missing area completion and component purification.

项目目录约定

Project Directory Convention

每次转换都创建独立项目目录：

text

bggg-creator-image2ppt/
└── projects/
    └── YYYYMMDD_slug/
        ├── original_inputs/
        ├── component_images/
        ├── imagegen_assets/
        ├── diagnostics/
        ├── manifest.json
        ├── output.pptx
        ├── summary.json
        └── process_notes.md

初始化：

bash

python3 bggg-creator-image2ppt/scripts/init_project.py pitch_deck \
  --source /path/to/reference.png \
  --date 20260504

Create an independent project directory for each conversion:

text

bggg-creator-image2ppt/
└── projects/
    └── YYYYMMDD_slug/
        ├── original_inputs/
        ├── component_images/
        ├── imagegen_assets/
        ├── diagnostics/
        ├── manifest.json
        ├── output.pptx
        ├── summary.json
        └── process_notes.md

Initialization:

bash

python3 bggg-creator-image2ppt/scripts/init_project.py pitch_deck \
  --source /path/to/reference.png \
  --date 20260504

二进制图片转 PPTX 工作流

Binary Image to PPTX Workflow

初始化项目，把源图复制到
```
original_inputs/
```
。
识别页面结构：
- 画布比例和大致尺寸。
- 背景是纯色、渐变、照片还是复杂插画。
- 每段文字的内容、位置、字号、颜色、粗细、对齐方式。
- 组件图片的边界、层级和是否需要透明背景。
默认用
```
imagegen
```
生成或编辑组件：
- 背景：完整画布、无文字、无前景组件。
- 照片/产品/人物/图标/复杂装饰：干净边缘，必要时透明背景。
- 图表：能原生重建就用形状和文本；复杂图表可先做成图片组件。
把组件图片放入
```
component_images/
```
或
```
imagegen_assets/
```
。
写 manifest，元素按从底到顶排序。

运行：

bash

python3 bggg-creator-image2ppt/scripts/image2pptx.py build \
  --manifest bggg-creator-image2ppt/projects/YYYYMMDD_slug/manifest.json \
  --output bggg-creator-image2ppt/projects/YYYYMMDD_slug/output.pptx \
  --summary bggg-creator-image2ppt/projects/YYYYMMDD_slug/summary.json

写
```
process_notes.md
```
，说明是否使用 imagegen、哪些对象是可编辑文本、哪些对象是图片 fallback。

Initialize the project and copy the source image to
```
original_inputs/
```
.
Identify page structure:
- Canvas ratio and approximate size.
- Whether the background is solid color, gradient, photo or complex illustration.
- Content, position, font size, color, weight, alignment of each text segment.
- Boundaries, hierarchy and whether transparent background is needed for component images.
Default to using
```
imagegen
```
to generate or edit components:
- Background: Complete canvas, no text, no foreground components.
- Photos/products/people/icons/complex decorations: Clean edges, transparent background if necessary.
- Charts: Reconstruct natively with shapes and text if possible; complex charts can be made into image components first.
Place component images into
```
component_images/
```
or
```
imagegen_assets/
```
.
Write the manifest, with elements sorted from bottom to top.

Run:

bash

python3 bggg-creator-image2ppt/scripts/image2pptx.py build \
  --manifest bggg-creator-image2ppt/projects/YYYYMMDD_slug/manifest.json \
  --output bggg-creator-image2ppt/projects/YYYYMMDD_slug/output.pptx \
  --summary bggg-creator-image2ppt/projects/YYYYMMDD_slug/summary.json

Write
```
process_notes.md
```
, indicating whether imagegen was used, which objects are editable text, and which objects are image fallbacks.

HTML/SVG 转 PPTX 工作流

HTML/SVG to PPTX Workflow

HTML/SVG 是代码形式的 PPT 时，优先走解析器：

bash

python3 bggg-creator-image2ppt/scripts/html_svg_to_manifest.py input.svg \
  --output bggg-creator-image2ppt/projects/YYYYMMDD_slug/manifest.json

python3 bggg-creator-image2ppt/scripts/html_svg_to_manifest.py input.html \
  --output bggg-creator-image2ppt/projects/YYYYMMDD_slug/manifest.json

解析原则：

原生还原
```
text
```
、
```
rect
```
、
```
ellipse
```
、
```
line
```
、简单图片。
HTML 里带绝对定位的
```
.slide
```
/
```
body
```
元素最容易被准确转换。
SVG 的复杂
```
path
```
、滤镜、渐变、mask、clip-path、foreignObject 可写入 diagnostics，并用 imagegen 或外部渲染结果作为图片组件 fallback。
解析脚本生成 manifest 后，仍用
```
image2pptx.py build
```
生成最终 PPTX。

When HTML/SVG is a code-form PPT, prioritize using the parser:

bash

python3 bggg-creator-image2ppt/scripts/html_svg_to_manifest.py input.svg \
  --output bggg-creator-image2ppt/projects/YYYYMMDD_slug/manifest.json

python3 bggg-creator-image2ppt/scripts/html_svg_to_manifest.py input.html \
  --output bggg-creator-image2ppt/projects/YYYYMMDD_slug/manifest.json

Parsing principles:

Natively restore
```
text
```
,
```
rect
```
,
```
ellipse
```
,
```
line
```
, and simple images.
```
.slide
```
/
```
body
```
elements with absolute positioning in HTML are easiest to convert accurately.
Complex SVG
```
path
```
, filters, gradients, mask, clip-path, foreignObject can be written into diagnostics, and use imagegen or external rendering results as image component fallbacks.
After the parsing script generates the manifest, still use
```
image2pptx.py build
```
to generate the final PPTX.

Manifest 格式

Manifest Format

manifest 是跨来源的中间层。坐标默认以源画布像素为单位，脚本会映射到 PowerPoint 尺寸。

json

{
  "deck": {
    "canvas_width": 1600,
    "canvas_height": 900,
    "slide_width_in": 13.333,
    "name": "Example Deck"
  },
  "slides": [
    {
      "name": "Cover",
      "elements": [
        {
          "kind": "background",
          "fill": "#f7f4ec"
        },
        {
          "kind": "image",
          "name": "Hero Product",
          "file": "component_images/product.png",
          "x": 910,
          "y": 170,
          "w": 520,
          "h": 520,
          "fit": "contain"
        },
        {
          "kind": "text",
          "name": "Title",
          "text": "Walk With Intention",
          "x": 120,
          "y": 180,
          "w": 680,
          "h": 150,
          "font_size_px": 68,
          "font_family": "Arial",
          "bold": true,
          "color": "#17120d",
          "align": "left"
        }
      ]
    }
  ]
}

常用元素：

```
background
```
:
```
fill
```
纯色或
```
file
```
背景图。

image

file

path

src

、

fit

。

fit

支持

stretch

、

contain

、

cover

。

text

text

、

font_size_px

或

font_size_pt

、

font_family

、

color

、

bold

、

italic

、

align

。

shape

shape

为

rect

、

roundRect

、

ellipse

或

line

；支持

fill

、

stroke

、

stroke_width_px

。

```
table
```
: 简单表格 fallback，用
```
rows
```
数组生成原生 PPT 表格。

The manifest is a cross-source intermediate layer. Coordinates are defaulted to source canvas pixels, and the script will map them to PowerPoint dimensions.

json

{
  "deck": {
    "canvas_width": 1600,
    "canvas_height": 900,
    "slide_width_in": 13.333,
    "name": "Example Deck"
  },
  "slides": [
    {
      "name": "Cover",
      "elements": [
        {
          "kind": "background",
          "fill": "#f7f4ec"
        },
        {
          "kind": "image",
          "name": "Hero Product",
          "file": "component_images/product.png",
          "x": 910,
          "y": 170,
          "w": 520,
          "h": 520,
          "fit": "contain"
        },
        {
          "kind": "text",
          "name": "Title",
          "text": "Walk With Intention",
          "x": 120,
          "y": 180,
          "w": 680,
          "h": 150,
          "font_size_px": 68,
          "font_family": "Arial",
          "bold": true,
          "color": "#17120d",
          "align": "left"
        }
      ]
    }
  ]
}

Common elements:

```
background
```
:
```
fill
```
for solid color or
```
file
```
for background image.

image

file

path

src

fit

fit

supports

stretch

contain

cover

text

text

font_size_px

font_size_pt

font_family

color

bold

italic

align

shape

shape

can be

rect

roundRect

ellipse

line

; supports

fill

stroke

stroke_width_px

```
table
```
: Simple table fallback, use
```
rows
```
array to generate native PPT tables.

可编辑性优先级

Editability Priority

文本框：标题、正文、页码、标签、按钮文字都优先转成原生 PPT 文本。
基础形状：矩形、圆角矩形、圆、线条优先转成 PPT 形状。
图片组件：照片、插画、图标、纹理、复杂图表作为独立图片层。
整页背景 fallback：无法拆干净时可保留一张底图，再把关键文字和组件覆盖为可编辑对象，并明确说明限制。

Text boxes: Titles, body text, page numbers, labels, button text are prioritized to be converted into native PPT text.
Basic shapes: Rectangles, rounded rectangles, circles, lines are prioritized to be converted into PPT shapes.
Image components: Photos, illustrations, icons, textures, complex charts as independent image layers.
Full-page background fallback: If it cannot be split cleanly, a base image can be retained, then key text and components are overlaid as editable objects, and limitations are clearly stated.

输出要求

Output Requirements

交付时至少说明：

项目目录路径。
PPTX 路径。
manifest 和 summary 路径。
可编辑文本框数量、图片组件数量、形状数量。
哪些组件由 Codex imagegen 生成或清理。
哪些复杂对象降级为图片 fallback。

At least explain the following when delivering:

Project directory path.
PPTX path.
Manifest and summary paths.
Number of editable text boxes, image components, shapes.
Which components were generated or cleaned by Codex imagegen.
Which complex objects were downgraded to image fallbacks.