prompt-engineer

Create production-ready prompts for Claude using a 5-phase methodology.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/walcon/opencode-example --skill prompt-engineer-walcon
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prompt-engineer
Source: https://github.com/walcon/opencode-example/tree/main/.opencode/skills/prompt-engineer
Command: npx skills add https://github.com/walcon/opencode-example --skill prompt-engineer-walcon

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill provides a disciplined 5-phase method to craft high-quality production prompts for Claude, including systematic requirements gathering, structure design, prompt construction, validation, and output generation. It reduces guesswork, speeds up prompt creation, and improves reliability.

Core Features & Use Cases

  • 5-phase methodology: Requirements gathering, Structure Design, Prompt Construction, Validation, Output Generation.
  • Pattern-driven design: Uses proven Prompt Patterns (PTF, COST, STAR, RRET, CoT, Few-Shot, ReAct, Constitutional) to cover a wide range of tasks.
  • Reusable templates: Scaffold prompts with role, instructions, output blocks and examples.
  • Use Case: Build a system prompt for a code reviewer or a data validator; generate a reusable template.

Quick Start

Save prompts to the prompts/<prompt-name>.md and validate with:

  • Validate format: npx tsx scripts/validate-prompt.ts prompts/my-prompt.md
  • Start from a simple template: copy prompts/summarize-article.md as a starting point.

Frequently Asked Questions about prompt-engineer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create production-ready prompts for Claude systematically?

Use the 5-phase methodology: requirements gathering, structure design, prompt construction, validation, and output generation. This structured approach reduces guesswork, ensures consistency with proven patterns like COST and ReAct, and produces reliable prompts faster than ad-hoc methods.

What prompt patterns should I use for different tasks?

Choose from PTF, COST, STAR, RRET, CoT, Few-Shot, ReAct, and Constitutional patterns. Each pattern addresses specific task types—use STAR for reasoning tasks, ReAct for multi-step workflows, Few-Shot for consistency, and Constitutional for value-aligned outputs.

Can I validate prompts before deployment?

Yes. Run `npx tsx scripts/validate-prompt.ts prompts/my-prompt.md` to check format compliance. Save prompts to `prompts/<prompt-name>.md` and validate structure, role definitions, instructions, output blocks, and examples for consistency.

How do I build a system prompt for an agent or specialized role?

Apply the 5-phase methodology with a role-based template scaffold. Define the agent's role, core instructions, output format blocks, and validation examples. Start from reusable templates like `summarize-article.md` and adapt for your specific agent function.

What's the difference between this approach and writing prompts ad-hoc?

Systematic methodology covers requirement gathering, design validation, and pattern alignment—eliminating inconsistency and trial-and-error. Ad-hoc prompts often miss edge cases; this approach ensures coverage across task requirements and proven reasoning patterns.

Can I reuse prompt templates across projects?

Yes. The Skill provides reusable templates with predefined role, instructions, output blocks, and examples. Templates scaffold new prompts quickly and maintain consistency across projects when adapted using the same 5-phase methodology.