writing-skills

Enforce test-driven RED-GREEN-REFACTOR compliance checks when authoring agent Skills.

Updated Feb 13, 2026
One-click install
npx skills add https://github.com/Andresch2/Workflows --skill writing-skills-andresch2
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: writing-skills
Source: https://github.com/Andresch2/Workflows/tree/main/ai-specs/skills/writing-skills
Command: npx skills add https://github.com/Andresch2/Workflows --skill writing-skills-andresch2

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill prevents agents from skipping or rationalizing around process rules when creating or updating documentation-based skills, especially under pressure.

Core Features & Use Cases

  • TDD for Skills: Uses RED-GREEN-REFACTOR so Skill content is validated by observing failures without the Skill, then enforcing compliance with the Skill.
  • Pressure-scenario testing: Guides creation of realistic multi-pressure scenarios to surface rationalizations agents use to bypass rules.
  • Loophole hardening: Adds explicit negations, rationalization tables, red flags, and re-verification until compliance holds.
  • CSO-ready discovery: Establishes discovery-focused guidance for writing a correct name/description that helps agents select the right Skill.
  • Documentation architecture: Recommends when to inline content vs reference other files, and how to organize long instructions.

Quick Start

Use writing-skills to edit or create a new Skill by running a failing baseline scenario first, then updating the Skill text to block the exact rationalizations you observed.

Frequently Asked Questions about writing-skills

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

FAQPage Schema
How do I stop AI agents from rationalizing and skipping process documentation rules under pressure?

To stop agents from skipping process documentation under pressure, apply test-driven authoring by running failing baseline scenarios, then harden instructions with explicit negations and rationalization tables. This enforces compliance by observing failures without the Skill first.

What is the test-driven approach for authoring agent Skills?

Test-driven Skill authoring is a process of running RED-GREEN-REFACTOR baseline checks to validate documentation rules. You observe agent failures without the Skill, then update the text to enforce compliance until the agent reliably follows instructions under pressure.

How do I harden agent instructions against loopholes and rationalization?

Harden agent instructions against loopholes by adding explicit negations, rationalization tables, and red flags. Re-verify compliance by testing the Skill against realistic multi-pressure scenarios until rationalizations are blocked.

How do I write Skill metadata for agent discovery and selection?

Write Skill metadata for agent discovery by establishing correct names and descriptions focused on CSO discovery requirements. This helps agents select the right Skill by defining clear, search-oriented guidance for the task it performs.

When should I use test-driven documentation for process rules?

Use test-driven documentation when creating, editing, or deploying Skill instructions, especially for discipline-enforcing rules agents may be tempted to bypass. It is needed when instructions must reliably hold under realistic multi-pressure scenarios.

How do I organize long Skill instructions and reference files?

Organize long Skill instructions by using documentation architecture to determine when to inline content versus reference other files. This structures complex instructions into modular components for better agent comprehension and compliance.