new-product
关于
This Claude Skill generates enterprise-grade product white papers by sourcing content from Google Drive documentation. It's triggered by requests like "create a product white paper" and is designed for SaaS, hardware, or service offerings. The skill follows a structured workflow to produce strategic, research-backed documents for executive and technical audiences.
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/new-product在 Claude Code 中复制并粘贴此命令以安装该技能
技能文档
Product White Paper
Generate enterprise-grade product white papers that serve as strategic assets for educating buyers and presenting research-backed solutions to complex business problems.
Workflow
Follow these steps to generate a product white paper:
-
Understand the Product: Clarify which product requires a white paper and identify the target audience (C-suite executives, technical decision-makers, or both)
-
Gather Source Material: Search Google Drive for existing product documentation, including:
- Product specifications and feature descriptions
- Technical architecture documentation
- Customer case studies or testimonials
- Competitive analysis
- ROI data or success metrics
- Implementation guides
- Governance and compliance documentation
-
Read the Structure Guide: Load
references/white-paper-essentials.mdto understand the required structure and tone -
Generate the White Paper: Create the white paper following the six-section structure, incorporating information from the gathered source material
-
Present in Chat: Output the complete white paper in the chat window formatted in markdown for easy copy-paste
Search Strategy
When searching Google Drive for source material, use multiple targeted searches:
For product information:
name contains '[product-name]' and (mimeType = 'application/vnd.google-apps.document' or mimeType = 'application/pdf')
For technical documentation:
fullText contains '[product-name]' and (fullText contains 'architecture' or fullText contains 'technical' or fullText contains 'specification')
For business value content:
fullText contains '[product-name]' and (fullText contains 'ROI' or fullText contains 'value' or fullText contains 'benefit' or fullText contains 'case study')
Cast a wide net initially, then narrow based on relevance. Review multiple documents to synthesize comprehensive content.
White Paper Structure
The white paper must follow this six-section structure (detailed in references/white-paper-essentials.md):
- Executive Summary - High-impact summary of strategic and technical takeaways
- The Strategic Imperative - Problem articulation, market context, and business justification
- The Solution Blueprint - Product features, capabilities, and operational benefits
- Building Trust - Governance, assurance, explainability, and risk mitigation
- Enterprise Value and ROI - Multi-dimensional value framework and workforce augmentation
- Implementation and Next Steps - Roadmap, change management, and clear call-to-action
Tone and Quality Standards
Adhere to these principles:
- Educational, not promotional: Establish expertise through depth and research, avoid overt sales language
- Formal and auditable: Use engineering/legal-grade formality with verifiable claims
- Research-backed: Ground all assertions in evidence from source materials
- Executive-relevant: Balance technical depth with strategic business value
- Conversion-oriented: Include clear takeaways and actionable next steps
Output Format
Present the white paper directly in the chat window using markdown formatting:
- Use proper heading hierarchy (# for title, ## for main sections, ### for subsections)
- Bold key terms and concepts for scannability
- Include bullet points for lists of features or benefits
- Maintain professional formatting for easy copy-paste into document editors
Key References
- Detailed structure and requirements: See
references/white-paper-essentials.mdfor comprehensive guidance on each section's purpose, required elements, and tone
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。
