Shenrui-Ma/video-regen-recipes / templates/drawing-process/svg-redraw-comparison
svg-redraw-comparison
Unverified用参考图制作高精度 SVG 临摹绘制过程视频,交付纯 SVG 网页、独立 SVG 和原图对照 MP4。禁止 Canvas、位图嵌入与外链,按构图、线稿、固有色、阴影、高光、细节重放;横图上下、竖图左右合成。
| File | Kind | Size |
|---|---|---|
| LICENSE | other | 1.1 KB |
| LICENSES.md | markdown | 998 B |
| README.md | markdown | 2.5 KB |
| SKILL.md | skill md | 3.8 KB |
| assets/manifest.json | config | 320 B |
| assets/reference.png | binary | 2.7 MB |
| examples/demo.html | other | 30.1 KB |
| examples/foundations.svg | other | 597.0 KB |
| examples/preview.mp4 | binary | 3.2 MB |
| examples/redraw.svg | other | 15.6 MB |
| examples/regions.json | config | 6.7 KB |
| examples/validation.json | config | 2.1 KB |
| profile.json | config | 2.2 KB |
| prompts/author-original.prompt.txt | markdown | 1.2 KB |
| prompts/redraw.prompt.txt | markdown | 1.2 KB |
| references/validation.md | markdown | 1.2 KB |
| references/workflow.md | markdown | 5.0 KB |
| scripts/_contours.py | script | 1.3 KB |
| scripts/_regions.py | script | 1.2 KB |
| scripts/_validate_svg.py | script | 1.5 KB |
| scripts/build_demo.py | script | 5.9 KB |
| scripts/clean_png.py | script | 2.1 KB |
| scripts/export_video.cjs | script | 6.1 KB |
| scripts/package_demo.py | script | 3.4 KB |
| scripts/trace_reference.py | script | 5.0 KB |
| sources.md | markdown | 1.7 KB |
| tests/check_player.cjs | script | 3.9 KB |
| tests/test_pipeline.py | script | 4.2 KB |
| web/load-art.js | script | 3.2 KB |
| web/player.css | other | 738 B |
| web/player.js | script | 2.3 KB |
sources.md
View on GitHub# 来源与验证
灵感来源:Bilibili **Hope麻匪**,[《再给 GPT6 Astra 一点时间》](https://www.bilibili.com/video/BV1azYW6zEXy)。已核对作者和视频标题;没有转载其视频或声称拿到了其SVG源码。
[作者提示词](prompts/author-original.prompt.txt)依据用户后来提供的清晰截图完整转录,替换了此前含乱码的转录。核心限制包括原生SVG、禁止Canvas/原图/Base64/外链、自包含HTML,以及有意义的绘制流程。
模板实现与适配:**Shenrui Ma(四倍体果蝇)**。
## 当前示例
参考图片由贡献者提供,已无损重编码、去元数据;1536×1024的解码RGBA像素与原件一致。图像本身不自动适用MIT代码许可。
当前SVG由连续色域轮廓矢量化与语义分层得到,包含22,798个最终色域路径,保留人物、服装、背景和光影。它不嵌入位图,也不能宣传成大模型逐笔手工绘画或思考过程录像。
浏览器按原尺寸渲染的平均每通道绝对误差为1.3002/255,PSNR为42.6144dB。仍有颜色量化误差,**不是严格逐像素1:1**;指标不等于对大模型能力的评测。
纯SVG Demo已检查Canvas/图片/外部脚本数量均为0,无外部资源请求。对比视频另由FFmpeg将原图与DOM截图合成,上下布局、40秒、24fps、960帧。视频实际导出并完整解码通过,见[验证记录](examples/validation.json)。
2026-09-12 将示例改为轻量 HTML 加本地 SVG 数据,HTML 小于 200 KB,完整色域路径与绘制阶段保持不变。离线选择加载、控件、回跳和五个关键时点的新旧像素对照检查通过;原 MP4 沿用此前产物,未重新编码。
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |