ai-native-productslisted

# AI-Native Product Development ## Purpose Provide patterns for building AI-powered product features from a PM perspective. Enable Claude to generate AI feature specifications, prompt engineering workflows, evaluation frameworks, and human-in-the-loop designs. Tailored for production AI research assistants and AI-augmented data products. ## When to Use - Designing AI-powered features or AI assistant capabilities - Writing prompt specifications for production AI features - Planning LLM integration architecture for a product - Building evaluation pipelines for AI output quality - Specifying human-in-the-loop review workflows - Creating A/B testing frameworks for AI features ## Examples **Example 1: AI feature specification** ``` User: "Spec out an AI-powered article summarizer for a crypto news platform" Claude: [Uses ai-native-products] # AI Article Summarizer - Feature Spec ## Behavior - Trigger: User clicks "Summarize" on any article - Output: 3-sentence summary + 3 key takeaways as bullet points - Latency target: <3 seconds (use streaming for perceived speed) - Source attribution: Summary must only use content from the article ## Prompt Specification System: "You are a crypto news analyst at a crypto media platform. Summarize articles for professional investors. Be precise with numbers and terminology." User: "<article>{article_text}</article> Provide: 1) A 3-sentence summary. 2) 3 key takeaways as bullets." ## Guardrails - [ ] No price predictions or directional