writing-skills

Apply the RED-GREEN-REFACTOR cycle to create and refine AI agent Skills.

1|Updated Nov 17, 2025
One-click install
npx skills add https://github.com/nimeshgurung/artifact-hub-collections --skill writing-skills-nimeshgurung
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: writing-skills
Source: https://github.com/nimeshgurung/artifact-hub-collections/tree/main/skills/raw/obra/superpowers/skills/writing-skills
Command: npx skills add https://github.com/nimeshgurung/artifact-hub-collections --skill writing-skills-nimeshgurung

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill ensures that documentation (Skills) is as rigorously tested and reliable as code, preventing ambiguity and ensuring agents consistently follow instructions.

Core Features & Use Cases

  • Test-Driven Skill Development: Applies the RED-GREEN-REFACTOR cycle to Skill creation and modification.
  • Bulletproofing Against Rationalization: Explicitly addresses and counters common excuses agents use to deviate from instructions.
  • Use Case: When creating a new Skill for code review, you would first write test scenarios to see how an agent fails without the Skill, then write the Skill to address those failures, and finally refactor to close any loopholes.

Quick Start

Use the writing-skills skill to create a new skill by first writing failing test scenarios.

Frequently Asked Questions about writing-skills

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

FAQPage Schema
How do I apply test-driven development to AI agent skill documentation?

Test-driven documentation applies the RED-GREEN-REFACTOR cycle to skill creation by first writing failing test scenarios with subagents to identify agent deviation, then writing documentation to pass those tests.

Why does my AI agent deviate from instructions and rationalize failures?

Agents deviate from instructions due to documentation ambiguity and rationalization loopholes. Identifying these failure modes through rigorous baseline testing before writing documentation ensures compliance and closes loopholes.

What is the RED-GREEN-REFACTOR cycle for skill development?

The RED-GREEN-REFACTOR cycle for skill development involves writing failing baseline tests to identify agent failure modes, creating documentation to make tests pass, and refactoring to close loopholes and ensure compliance.

How do I bulletproof AI agent instructions against rationalization?

Bulletproofing against rationalization involves identifying agent deviation through baseline subagent testing, then iteratively refactoring skill documentation to explicitly address and counter common excuses agents use to ignore instructions.

Do I need baseline subagent testing before writing skill documentation?

Yes, baseline subagent testing is required to identify failure modes and documentation ambiguity before writing skill documentation, ensuring the final instructions are rigorously tested and reliable.