AI Prompt Engineering

Structured Instructions & AI Workflows

I explore prompt engineering to guide AI tools in software development, content creation, automation, and problem-solving. I focus on writing structured instructions and improving workflows to get more useful outputs from AI systems.

Development Prompt (Demonstration)

Component Refactoring & Modularization

Used to transform unstructured vanilla JavaScript into clean, event-driven module structures.

Act as an expert frontend engineer. Review the provided code snippet. Refactor it into a self-contained ES6 class or module pattern. Ensure zero global namespace pollution, attach clean event delegation, and include TypeScript JSDoc type annotations.

Website-Building Prompt (Demonstration)

Accessible Semantic Structure Generation

Directs AI models to output fully accessible semantic HTML5 with strict ARIA standards.

Generate semantic HTML5 markup for an e-commerce product card. Requirements: Proper heading hierarchy, accessible button states, alt attributes for product imagery, focus-visible indicators, and schema.org Product microdata.

Content Generation (Demonstration)

Product Story Formulation

Generates authentic, uninflated product descriptions for physical apparel.

You are a senior brand writer. Craft a 60-word description for an everyday cotton hoodie. Focus on tactile comfort, durability, and daily utility. Do not use hyperbolic marketing cliches like 'revolutionary' or 'game-changer'. Keep the tone warm, grounded, and human.

Automation Workflow (Demonstration)

JARVIS Assistant Intent Parser

System instructions for mapping voice transcripts to discrete automation commands.

You are a local voice command parser. Given user voice transcript input, extract intent and parameters into valid JSON: {"action": "open_app"|"create_note"|"query", "target": string, "confidence": float}. Output only pure JSON without markdown wrap.