construct-skill

Authors new SKILL.md files through a TDD-style RED-GREEN-REFACTOR loop.

2|Updated Apr 4, 2026
One-click install
npx skills add https://github.com/DoctorMozg/claude-pipelines --skill construct-skill-doctormozg
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: construct-skill
Source: https://github.com/DoctorMozg/claude-pipelines/tree/main/plugins/mz-dev-base/skills/construct-skill
Command: npx skills add https://github.com/DoctorMozg/claude-pipelines --skill construct-skill-doctormozg

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? New skills often ship with weak descriptions, missing sections, and no defense against the rationalizations that cause future runs to skip them. This Skill turns the SKILL_GUIDELINES.md rulebook into an executable authoring workflow so every new skill is built against the excuses it will face. ## Core Features & Use Cases - TDD Authoring Loop: RED enumerates rationalizations the skill will face, GREEN writes the canonical 7-section anatomy that rebuts them, REFACTOR stress-tests the draft as a skeptic. - Skill-Type Classification: Classifies each new skill as discipline, collaboration, or reference and applies the matching persuasion register (Authority/Commitment/Social-Proof vs Unity vs neutral). - Pre-Publish Checklist & Approval Gate: Runs a 20-bullet compliance checklist and presents the full draft for explicit user approval before saving. - Use Case: When you ask to "add a retry-policy skill to mz-dev-pipe", this Skill walks you through rationalization discovery, anatomy drafting, and checklist verification before writing the final SKILL.md. ## Quick Start Ask the assistant to write a new skill named retry-policy for the mz-dev-pipe plugin and let it guide you through the authoring loop.

Frequently Asked Questions about construct-skill

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

FAQPage Schema
How do I write a new skill for a Claude Code plugin?

Invoke this Skill with the skill name and target plugin. It walks you through RED (enumerate rationalizations), GREEN (write the 7-section anatomy), and REFACTOR (stress-test the draft), ending with an approval gate before the SKILL.md is saved.

What sections must a SKILL.md contain?

The canonical anatomy requires seven sections in order: Overview, When to Use, Core Process, Techniques, Common Rationalizations, Red Flags, and Verification. Sections that genuinely do not apply must be marked N/A with a rationale rather than omitted.

What makes a skill description CSO-compliant?

A compliant description leads with ALWAYS invoke when, lists 2-3 concrete trigger phrases, stays under 250 characters, uses third person, and contains no workflow summary. It acts as the skill's auction bid for invocation.

When should I not use this skill authoring workflow?

Do not use it to audit or fix already-merged skills, to edit phase files without changing SKILL.md, or to write non-skill markdown like rules or plan docs. Use /review-branch for compliance audits of existing skills.

Why do discipline skills need an anti-rationalization table?

Discipline skills push back against shortcuts, so they must pre-empt the excuses future runs will raise. The table requires at least 3 verbatim rationalizations paired with specific rebuttals citing concrete failure modes, never generic best-practice claims.