What is the purpose of writing-skills?

To apply TDD to skill creation and verification by writing pressure tests and SKILL.md to ensure correct agent behavior and close loopholes.

What should frontmatter contain?

Only name and description should be present in YAML frontmatter.

How does this improve deployment quality?

It enforces a failure-before-success test pattern, ensuring the skill actually teaches the right behavior and is free of loopholes before production use.

writing-skills

Scanned

npx machina-cli add skill a5c-ai/babysitter/writing-skills --openclaw

Files (1)

SKILL.md

1.1 KB

Writing Skills

Overview

Writing skills IS Test-Driven Development applied to process documentation. Write pressure tests, watch agents fail, write the skill, watch them pass, close loopholes.

Core principle: If you did not watch an agent fail without the skill, you do not know if the skill teaches the right thing.

TDD for Skills

TDD Concept	Skill Creation
Test case	Pressure scenario with subagent
Production code	Skill document (SKILL.md)
RED	Agent violates rule without skill
GREEN	Agent complies with skill present
REFACTOR	Close loopholes

Skill Structure

YAML frontmatter: name and description only
Description: "Use when..." (triggering conditions only, never summarize workflow)
Flat namespace, separate files only for heavy reference or reusable tools

Tool Use

Meta-skill for creating new skills within the methodology.

Source

git clone https://github.com/a5c-ai/babysitter/blob/main/plugins/babysitter/skills/babysit/process/methodologies/superpowers/skills/writing-skills/SKILL.md

View on GitHub

Overview

Writing skills apply Test-Driven Development to process documentation. You write pressure tests, observe agent failures, craft the skill, then validate that agents pass with the skill and fail without it, closing loopholes. The core principle is that you must witness failure without the skill to prove the skill teaches the right behavior.

How This Skill Works

First, write a pressure test as a subagent scenario and align it with a SKILL.md production artifact. Use RED-GREEN-REFACTOR to verify: without the skill, the agent breaks the rule; with the skill, the agent complies, and loopholes are closed through refactoring.

When to Use It

When creating a new skill
When editing an existing skill
Before deploying a skill to production to verify behavior
To close loopholes by testing edge cases
During regression testing after updates to the skill or related processes

Quick Start

Step 1: Write a pressure test scenario and create the SKILL.md frontmatter with name and description
Step 2: Run the RED-GREEN-REFACTOR cycle to verify failure without the skill and success with it
Step 3: Validate readiness for deployment and record verification results

Best Practices

Write a pressure test scenario before drafting SKILL.md
Keep YAML frontmatter to only name and description
Validate the red, green, and refactor cycles with real agent behavior
Iterate to close loopholes and edge cases
Document the justification and verification criteria for the skill

Example Use Cases

Create a new skill to enforce a data handling policy and validate with a pressure test
Edit an existing skill to tighten its triggering conditions
Run a pre-deployment test to ensure agents violate rules without the skill and comply with it
Refactor after testing to close loopholes
Document a complex skill with subagents and reusable tools using the TDD approach

Frequently Asked Questions

Add this skill to your agents