What are common prompt anti-patterns this skill fixes?

It eliminates prescriptive 'how-to' steps, arbitrary limits, weak language like 'try to', and buried critical information that reduces model performance.

When should I use the Prompt Engineering skill?

Use it whenever you are creating new LLM prompts, reviewing existing prompt structures for clarity, or building specialized Claude Code skills.

Does this skill support specific prompt structures?

Yes, it provides standardized templates for Skills, Agents, and System Instructions, ensuring consistent and professional architecture across your AI projects.

How does this skill help with long AI workflows?

It utilizes the Memento pattern to manage context rot and limited working memory by logging findings after each step and refreshing context before final synthesis.

What is first-principles prompt engineering?

It is an approach that focuses on defining clear goals (the WHAT) and constraints (the WHY) while trusting the model's inherent capabilities to determine the best execution path (the HOW).

Prompt Engineering Architect

Name: Prompt Engineering Architect
Author: doodledood

bydoodledood

•

数据科学与机器学习

Crafts and optimizes high-performance LLM prompts using first-principles thinking and rigorous context discovery.

The Prompt Engineering skill transforms how you interact with AI by moving beyond trial-and-error implementation. It guides you through a structured discovery process to surface domain knowledge and constraints before generating prompts that define clear goals and success criteria without over-prescribing steps. Whether you are creating new system instructions, updating existing skills, or implementing complex multi-phase workflows with the Memento pattern, this skill ensures your prompts are dense, unambiguous, and optimized for model performance.

主要功能

01Standardized templates for creating Skills, Agents, and System Instructions.

0263 GitHub stars

03Implementation of the Memento pattern for managing context in complex multi-phase workflows.

04First-principles prompt design focusing on goals and constraints rather than procedural steps.

05Structured context discovery interviews to surface latent requirements and domain-specific jargon.

06Automated identification and removal of anti-patterns like weak language and arbitrary limits.

使用场景

01Refactoring bloated system prompts to improve information density and model adherence.

02Designing complex data extraction pipelines that require state tracking across multiple steps.

03Building specialized Claude Code skills with clear missions and validation checklists.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add doodledood/manifest-dev prompt-engineering

For use in Claude.ai and ChatGPT

主要功能

01Standardized templates for creating Skills, Agents, and System Instructions.

0263 GitHub stars

03Implementation of the Memento pattern for managing context in complex multi-phase workflows.

04First-principles prompt design focusing on goals and constraints rather than procedural steps.

05Structured context discovery interviews to surface latent requirements and domain-specific jargon.

06Automated identification and removal of anti-patterns like weak language and arbitrary limits.

使用场景

01Refactoring bloated system prompts to improve information density and model adherence.

02Designing complex data extraction pipelines that require state tracking across multiple steps.

03Building specialized Claude Code skills with clear missions and validation checklists.