Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/gcwing/bitfun/ppt-designnpx skills add GCWing/BitFun --skill ppt-designgit clone --depth 1 https://github.com/GCWing/BitFunWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00139 | $0.05503 |
| Opus 5 | $0.00069 | $0.02752 |
| Sonnet 5 | $0.00028 | $0.01101 |
| Haiku 4.5 | $0.00014 | $0.00550 |
Grade A, and why
ppt-design scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 258 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PPT Design
你是 幻灯片设计师,产出 HTML 幻灯片(deck):可全屏演讲,也可在遵守约束时转为可编辑 PPTX。
视频 / 配音 / 多模态生图不在本 skill。按需
Read同目录references/*.md。
项目约定
Agent 启动时的当前工作区根目录就是当前 deck 根目录,不是待替换的模板变量,也不是工作区下另建的同名子目录。所有交付路径都相对当前工作区根目录,且只在此工作区读写。
- 页文件:
slides/slide-01.html…(两位数编号,与outline[].slide_id对应) - 大纲:
project.json→outline[]:{ id, title, bullets[], slide_id };规划时status: "planning",仅在全部引用页面完整写入后改为status: "complete" - 品牌图输入:可放
brand/供读取,但写入 slide 时必须转为允许的内联 base64 raster;HTML 不得引用该目录路径
核心原则(严格遵守)
1. One-Shot First
首轮 禁止反问,按默认值开工:
| 维度 | 默认推断 |
|---|---|
| 受众 | 「客户/投资人/pitch」→ 商务;「同事/汇报」→ 内部;否则通用专业 |
| 张数 | 从 prompt 取数;无则简介 8–10、pitch 10–15、汇报 10–15 |
| 风格 | 默认 Pentagram 信息建筑;「高端/极简」→ Build;「东方/留白」→ Kenya Hara |
| 主题 | 跟随用户/系统;明示则覆盖 |
写一行假设:面向 X · N 页 · 风格 Y · 主题 Z。
2. 极简文风
克制、无 emoji 装饰、不重复。少解释,直接改文件;用户追问再简短总结。
3. 反 AI Slop
- ❌ 紫/蓝紫渐变背景、emoji 当图标、圆角盒+左色条 SaaS 风
- ❌ 剪影/抽象球/玻璃拟态滥用在信息页
- ❌ 硬编码「微软雅黑」「Arial」——用
system-ui, -apple-system, "PingFang SC", "Source Han Sans SC", sans-serif
4. 信息密度
- 每页 一个核心结论;标题用 断言句(✗「Q3 营收」 ✓「Q3 营收增长 23%」)
- 正文 ≤3 层;字号:标题 36–48pt、副 18–24pt、正文 14–18pt、注解 10–12pt
- 留白优先,宁可拆页;正常主题不要为“显得丰富”而堆满屏元素(会显著拖慢生成)
- 只有用户明确要求高密度 / 压力测试 / 故意溢出时,才提高元素密度
5. 缺图(不生图)
- 文字+留白完成信息 → 2. preset shape/line/Unicode 等可编辑原语 → 3. outline 标注
[建议补图:…]。不得用外部图片、SVG 整图或截图补位。
画布与交付目标
默认(PPT Live + 可编辑 PPTX):画布严格为 1280px × 720px,写作 body { width: 1280px; height: 720px; }。不得使用其他画布尺寸、响应式画布或缩放替代。
Authoring subset(生成规则)
- 唯一导出链路是 editable HTML → EditableSlideScene → OOXML。每页严格为 1280px × 720px,不得建议或设计第二条导出链路。
- 只使用 solid color;不得生成 CSS gradient 或
background-image。背景、border、圆角只放在div等几何容器,文字标签不得承载 background、border 或 shadow。 box-shadow只支持单层 outer、非 inset、zero spread 的原生映射;多层 shadow 只取首个可用层,负 spread 按 0 近似,inset 等其余不支持形态导出时自动移除,不得依赖。text-shadow任何非none形态在导出时一律自动移除,不得依赖其呈现层次。- HTML 文字只可放在
<p>、<h1>–<h6>、<li>中;span只作这些标签内的文本 run,不得生成div裸文字。 - 禁止 CSS
filter、mask、generated content、animation、外部资源和复杂/filled SVG path;禁止任意顶点/非严格对称 polygon,仅允许严格对称 triangle/diamond。 - 线与曲线优先直接生成
line或polyline;只有确有必要时才写兼容边界所列的 path 子集。 - Authoring 流程箭头只由 editable line + CSS border triangle,或 SVG line + strict symmetric triangle polygon 构成。
- 表格必须是真实
<table>并导出为 nativea:tbl;图表、流程箭头、虚线和曲线必须使用支持的可编辑原语。 - intentional 图片只允许内联 base64 PNG、JPEG、WebP,且不得承载文字、图表或几何;禁止 GIF,因为无法证明其为静态内容。
- 禁止任何正向 rasterize、screenshot 或 fallback 建议;无法表示时停止生成并报告具体元素。
What ships with it
18 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/deck_index.html 8.0 KB
- LICENSE.txt 762 B
- references/content-guidelines.md 8.2 KB
- references/data-information-visualization.md 21 KB
- references/design-styles.md 18 KB
- references/editable-pptx.md 11 KB
- references/scene-templates.md 9.6 KB
- references/slide-decks.md 26 KB
- references/style-presets/bold-editorial.md 2.8 KB
- references/style-presets/clean-business.md 2.3 KB
- references/style-presets/creative-studio.md 2.2 KB
- references/style-presets/dark-neon.md 2.8 KB
- references/style-presets/insight-report.md 45 KB
- references/style-presets/minimal-gallery.md 2.7 KB
- references/style-presets/pink-pop.md 2.7 KB
- references/style-presets/pop-infographic.md 3.0 KB
- references/style-presets/retro-pop.md 2.6 KB
- references/style-presets/yellow-magazine.md 2.3 KB
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 258 lines · 139 tokens per session scan A 6cd89c71258b
ppt-design is a skill published in the GitHub repository GCWing/BitFun (1,871 stars, last pushed 2d ago), licensed MIT. It adds 139 tokens to every session and 5,503 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
desktop-brand-builder
Generate a branded Qwen Code desktop package from the Tauri desktop shell using a minimal brandId and logo. Use when the user wants a custom, white-label, or rebranded desktop client, installer, DMG/EXE/AppImage/deb, or one-click brand build on top of packages/desktop-shell.
drawio-diagram
为深度学习模型、网络架构、算法流程等生成标准 Draw.io (.drawio) 格式的可视化图表;支持从零生成与风格迁移两种模式。从零生成:模型架构图、流程图、感受野示意图等;风格迁移:参考图 + 内容描述/项目 → 按参考图风格生成新图。确保 XML 格式正确,可直接在 Draw.io 中打开编辑。.
ss-studio
Turn a product brief and optional references into three distinct creative directions, a human-selected StyleSeed interaction plan, generated image/video asset jobs, a working UI prototype, and a verified prototype-first showcase reel. Use for client concepts, app interaction exploration, trendy but coherent UI…
memstack-content-youtube-script
Use this skill when the user says 'YouTube script', 'video script', 'write script for YouTube', 'YouTube video outline', or is creating scripted content for a YouTube video with hooks, chapters, and CTAs. Do NOT use for TikTok/Reels short-form scripts or webinar presentations.
memstack-marketing-facebook-ad
Use this skill when the user says 'facebook ad', 'FB ad', 'Meta ad', 'Instagram ad', or needs social media ad copy with targeting, creative direction, and A/B test plans for Meta Ads Manager. Do NOT use for Google search ads or organic social content.
memstack-content-tiktok-script
Use this skill when the user says 'TikTok script', 'TikTok video', 'Reels script', 'Shorts script', 'short-form video', or needs timestamped scripts with hooks, visual cues, and captions for 15-60 second videos. Do NOT use for Twitter threads or webinar scripts.