writing-skills

Apply test-driven development cycles to create and refine process documentation.

130|8|Updated Jan 23, 2026
One-click install
npx skills add https://github.com/sandgardenhq/sgai --skill writing-skills-sandgardenhq
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: writing-skills
Source: https://github.com/sandgardenhq/sgai/tree/main/cmd/sgai/skel/.sgai/skills/meta/writing-skills
Command: npx skills add https://github.com/sandgardenhq/sgai --skill writing-skills-sandgardenhq

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Writing or editing process documentation often leads to unclear, unused, or loophole-ridden skills that agents ignore under pressure. This Skill applies Test-Driven Development to documentation so you create concise, discoverable, and enforceable skills that actually change behavior.

Core Features & Use Cases

  • TDD for Documentation: RED-GREEN-REFACTOR cycle using subagents to create failing baselines, write minimal guidance, and close rationalization loopholes.
  • Discovery & Token Efficiency: Rich, symptom-focused descriptions, keyword strategy, naming conventions, and strict brevity to keep frequently-loaded skills lightweight.
  • Testing Playbooks: Tailored test strategies for discipline, technique, pattern, and reference skills, plus checklists, rationalization tables, and red-flag heuristics.
  • Use Case: Before adding a new “condition-based-waiting” technique, you run subagent pressure scenarios to observe failures, write the minimal SKILL.md addressing those gaps, and iterate until agents consistently comply.

Quick Start

Ask the assistant to run a baseline subagent scenario without any skill, capture the failures and rationalizations, then write the minimal SKILL.md and re-test until it passes under pressure.

Frequently Asked Questions about writing-skills

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

FAQPage Schema
How do I test process documentation to ensure AI agents follow it under pressure?

Test process documentation by running subagent pressure scenarios to capture baseline failures and rationalizations, then iterate on the SKILL.md until agents consistently comply under pressure.

What is test-driven development for documentation and how does it work?

Test-driven development for documentation applies a RED-GREEN-REFACTOR cycle using subagents to create failing baselines, write minimal guidance, and close rationalization loopholes so skills remain enforceable.

How do I make AI skills more discoverable and token-efficient?

Make AI skills discoverable and token-efficient by writing symptom-focused descriptions, applying keyword strategies, following naming conventions, and enforcing strict brevity for frequently-loaded skills.

What's the best way to standardize emerging patterns and tools across multiple projects?

Standardize emerging patterns across projects by applying test-driven documentation cycles, verifying effectiveness with subagents, and enforcing YAML-frontmatter structure and naming conventions.

Why does my AI agent ignore process documentation when executing complex tasks?

AI agents ignore process documentation when loopholes exist; you can close rationalization gaps by writing minimal guidance and testing with subagent pressure scenarios until compliance is consistent.

Do I need subagents to create and verify effective process documentation?

Yes, subagents are required to execute the test-driven documentation cycle, capture baseline failures, verify guidance effectiveness, and close rationalization loopholes under pressure scenarios.