presentation-outline
关于
This skill converts a Google Doc into a structured 10-slide presentation outline with titles, subtitles, and bullet points. It's ideal for developers needing to quickly draft a presentation from existing technical documentation. The output is provided directly in chat for immediate review and further editing.
快速安装
Claude Code
推荐/plugin add https://github.com/christopheryeo/claude-skillsgit clone https://github.com/christopheryeo/claude-skills.git ~/.claude/skills/presentation-outline在 Claude Code 中复制并粘贴此命令以安装该技能
技能文档
Presentation Outline
Generate a structured presentation outline from a Google Doc to quickly transform topic documentation into a presentation-ready format.
How It Works
This skill takes a Google Doc containing information about a topic and transforms it into a presentation outline with the following structure per slide:
- Slide Title - The main topic or concept
- Slide Subtitle - Context or focus area (can be optional)
- Bullet Points - Supporting details (one subject per slide, 3-5 bullets)
The output is limited to a maximum of 10 slides for an optimal presentation length.
Process
- Provide a Google Doc - Share the URL or ID of a Google Doc containing the topic information
- Outline Generation - Claude extracts key concepts and organizes them into distinct slides
- Chat Display - The outline is displayed in the chat window for you to review and edit
- Refinement - Adjust titles, subtitles, or bullets as needed
Guidelines for Best Results
For detailed best practices on what makes an effective presentation outline, see references/guide.md. Key principles include:
- One topic per slide - Each slide focuses on a single main concept
- Concise titles - 3-8 words that clearly identify the slide topic
- Supporting bullets - 3-5 bullet points per slide that support the main topic
- Optimal length - 8-10 slides provides balanced pacing for a 20-30 minute presentation
- Logical flow - Clear progression from opening to conclusion
Output Format
The outline is displayed in markdown format in the chat with clear delineation between slides:
# Slide 1: [Title]
**Subtitle**: [Subtitle if applicable]
- Bullet point 1
- Bullet point 2
- Bullet point 3
# Slide 2: [Title]
**Subtitle**: [Subtitle if applicable]
- Bullet point 1
- Bullet point 2
- Bullet point 3
Tips
- Longer, more detailed source documents produce better outlines
- Ensure your Google Doc clearly separates different topics and concepts
- Review the generated outline and edit slide titles/bullets for your specific audience
- If the outline exceeds 10 slides, consider combining related slides or removing less critical content
GitHub 仓库
相关推荐技能
content-collections
元Content Collections 是一个 TypeScript 优先的构建工具,可将本地 Markdown/MDX 文件转换为类型安全的数据集合。它专为构建博客、文档站和内容密集型 Vite+React 应用而设计,提供基于 Zod 的自动模式验证。该工具涵盖从 Vite 插件配置、MDX 编译到生产环境部署的完整工作流。
sglang
元SGLang是一个专为LLM设计的高性能推理框架,特别适用于需要结构化输出的场景。它通过RadixAttention前缀缓存技术,在处理JSON、正则表达式、工具调用等具有重复前缀的复杂工作流时,能实现极速生成。如果你正在构建智能体或多轮对话系统,并追求远超vLLM的推理性能,SGLang是理想选择。
evaluating-llms-harness
测试该Skill通过60+个学术基准测试(如MMLU、GSM8K等)评估大语言模型质量,适用于模型对比、学术研究及训练进度追踪。它支持HuggingFace、vLLM和API接口,被EleutherAI等行业领先机构广泛采用。开发者可通过简单命令行快速对模型进行多任务批量评估。
llamaguard
其他LlamaGuard是Meta推出的7-8B参数内容审核模型,专门用于过滤LLM的输入和输出内容。它能检测六大安全风险类别(暴力/仇恨、性内容、武器、违禁品、自残、犯罪计划),准确率达94-95%。开发者可通过HuggingFace、vLLM或Sagemaker快速部署,并能与NeMo Guardrails集成实现自动化安全防护。
