skills.rest
HalseyYang
HalseyYangCOMMUNITY
Explore all skills by this author →
GitHub

writing-skills

Create and test agent skills using test-driven development methodology.

1|Updated Jul 13, 2026
Search Tags:#documentation#prompt engineering#claude skills#subagents#test-driven development#skill authoring#agent testing
One-click install
npx skills add https://github.com/HalseyYang/Skills-HY --skill writing-skills-halseyyang
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: writing-skills
Source: https://github.com/HalseyYang/Skills-HY/tree/main/writing-skills
Command: npx skills add https://github.com/HalseyYang/Skills-HY --skill writing-skills-halseyyang

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires graphviz.

What problem does it solve? Writing agent skills without testing them leads to documentation that agents ignore, misinterpret, or rationalize away under pressure. This Skill applies test-driven development (RED-GREEN-REFACTOR) to skill authoring so every skill is verified against real agent behavior before deployment. ## Core Features & Use Cases - TDD-Based Skill Authoring: Run baseline pressure scenarios with subagents (RED), write the minimal skill addressing observed failures (GREEN), then close rationalization loopholes (REFACTOR). - Discovery Optimization (CSO): Guidelines for writing descriptions, keywords, and names so future agents actually find and load the skill, including the rule that descriptions must state only triggering conditions, never workflow summaries. - Pressure Testing Methodology: Reference material on building multi-pressure scenarios (time, sunk cost, authority, exhaustion) and persuasion principles grounded in published research. - Use Case: You want to add a new discipline-enforcing skill (e.g., mandatory code review) to your agent library. Use this Skill to first watch a subagent violate the rule without the skill, then author and bulletproof the skill until the agent complies under maximum pressure. ## Quick Start Use the writing-skills skill to create a new agent skill for my preferred code review workflow, testing it with subagent pressure scenarios before deployment.

Frequently Asked Questions about writing-skills

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

FAQPage Schema
How do I write an effective SKILL.md file for an AI agent?▼

Start with YAML frontmatter containing name and description, where the description begins with "Use when..." and lists only triggering conditions. Follow with an overview, when-to-use section, quick reference table, and one excellent code example, keeping the body under 500 lines.

How do I test an agent skill before deploying it?▼

Run pressure scenarios with subagents without the skill first to capture baseline failures and verbatim rationalizations. Then write the skill addressing those failures, re-test with the skill present, and iterate to close any new loopholes agents discover.

What should go in a skill description field?▼

The description should state only when to use the skill, written in third person starting with "Use when..." plus specific symptoms and triggers. Never summarize the skill's workflow, because agents may follow the description shortcut instead of reading the full skill body.

When should I not create a new agent skill?▼

Skip skill creation for one-off solutions, project-specific conventions better placed in CLAUDE.md, well-documented standard practices, and mechanical constraints that can be enforced with regex or validation scripts instead of documentation.

Why do agents ignore or bypass skill instructions under pressure?▼

Agents rationalize violations when facing combined pressures like time limits, sunk cost, and authority. Counter this with explicit negations of specific excuses, rationalization tables, red-flag lists, and foundational principles like "violating the letter is violating the spirit."

Related Skills

Explore More Software Engineering Skills→
growilabsgrowilabs
1.5k

kiro-debug

Extract error messages and traces to classify root causes and produce fix plans.

Official
Advanced
ronniegeraghtyronniegeraghty
3

reskill

Audit agent charters and histories to extract reusable patterns into team skills.

Community
Advanced
sswapnil2sswapnil2

test-driven-development

Enforces test-first development with RED-GREEN-REFACTOR cycles for fintech projects.

Community
Advanced
yersonargotevyersonargotev
1

principle-encode-lessons-in-structure

Convert recurring corrections into lint rules, metadata flags, runtime checks, or scripts.

Community
Basic