writing-skills

Create and edit skills using test-driven development with pressure testing.

115|13|Updated Mar 24, 2026
One-click install
npx skills add https://github.com/Houseofmvps/ultraship --skill writing-skills-houseofmvps
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: writing-skills
Source: https://github.com/Houseofmvps/ultraship/tree/main/skills/writing-skills
Command: npx skills add https://github.com/Houseofmvps/ultraship --skill writing-skills-houseofmvps

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the critical problem of creating ineffective or broken skills that agents cannot discover, follow, or trust under pressure. Without proper test-driven development and structure, skills often fail when needed most, leading to repeated mistakes and wasted effort.

Core Features & Use Cases

  • TDD for Documentation: Applies test-driven development principles to skill creation, ensuring skills are tested with pressure scenarios before deployment.
  • Bulletproof Design: Teaches how to structure skills to resist rationalization and ensure compliance even under time pressure, sunk cost, and authority challenges.
  • Discovery Optimization: Provides Claude Search Optimization techniques to ensure skills are found and loaded when relevant.
  • Use Case: When you need to create a new skill for Claude Code or other AI agents, this Skill guides you through baseline testing, minimal documentation, and loophole closure.

Quick Start

Use the writing-skills skill to create a new skill by first running a baseline pressure scenario without the skill, then writing the skill to address the observed failures.

Frequently Asked Questions about writing-skills

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

FAQPage Schema
How do I create AI agent skills that remain reliable under pressure scenarios?

Test-driven development for process documentation ensures AI agent skills resist rationalization by running baseline tests, writing minimal documentation, and applying subagent pressure testing to close loopholes.

What is the RED-GREEN-REFACTOR cycle for test-driven documentation?

The RED-GREEN-REFACTOR cycle applies test-driven development to documentation by first observing baseline failures, writing minimal skill instructions to pass, and refining to ensure agent compliance.

How do I write bulletproof Claude Code skills that agents follow correctly?

Bulletproof Claude Code skills require testing against authority challenges and sunk cost scenarios, applying discovery optimization techniques, and verifying subagent compliance to prevent rationalization.

Can I use test-driven development to edit existing skills and verify their compliance?

Yes, test-driven development methodology applies to editing existing skills and verifying compliance by running pressure scenarios that test whether agents correctly follow the updated documentation.

Why do AI agent skills fail when needed most and how can I prevent it?

AI agent skills fail under pressure due to improper structure and lack of loophole closure, which test-driven documentation prevents through baseline testing and subagent pressure verification.

Do I need to run a baseline test before writing a new skill for Claude Code?

Yes, running a baseline pressure scenario without the skill reveals observed failures, providing the necessary context to write effective documentation that addresses actual agent compliance gaps.