skills.rest
ysuurme
ysuurmeCOMMUNITY
Explore all skills by this author →
GitHub

write-skills

Convert failing baseline rationalizations into closed loopholes within SKILL.md.

Updated Mar 2, 2022
Search Tags:#skill design#MECE#agent workflows#documentation templates#routing signals#baseline-first#prompt constraints
One-click install
npx skills add https://github.com/ysuurme/azure_hello_world --skill write-skills-ysuurme
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: write-skills
Source: https://github.com/ysuurme/azure_hello_world/tree/main/.agents/skills/write-skills
Command: npx skills add https://github.com/ysuurme/azure_hello_world --skill write-skills-ysuurme

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents AI agents from using inconsistent, over-scoped, or poorly routed skills by enforcing a baseline-first, MECE-compliant drafting process for every skill update.

Core Features & Use Cases

  • Skill creation and updates with guardrails: guides writing and editing of SKILL.md/REFERENCE.md/EXAMPLES.md so the skill is scannable, testable, and scoped correctly.
  • Baseline-first constraint writing: turns documented rationalizations from a failing baseline into closed loopholes that the skill forbids.
  • MECE enforcement and split detection: checks the AGENTS.md index and adjacent skills’ scope to avoid overlap, gaps, and trigger overlap, and signals when splitting is required.
  • Routing-signal correctness: ensures every exclusion and mandatory pre-read includes a valid → skill-name pointer so agents can route without dead ends.

Quick Start

Use the write-skills skill to create or revise a skill by first running a failing pressure scenario, extracting the agent’s exact rationalizations, and then writing constraints that explicitly forbid those rationalizations.

Frequently Asked Questions about write-skills

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

FAQPage Schema
How do I write agent skills that are MECE-compliant and properly scoped?▼

To write MECE-compliant agent skills, enforce a baseline-first drafting workflow that partitions scope into Owns, Does not own, and Interfaces with, preventing overlap and gaps across your agent directory.

What is the baseline-first approach for writing prompt constraints in SKILL.md?▼

The baseline-first approach runs a failing pressure scenario, extracts the agent's exact rationalizations, and writes constraints that explicitly forbid those rationalizations, closing loopholes in your SKILL.md.

How do I prevent routing dead ends when configuring agent skill exclusions?▼

To prevent routing dead ends, ensure every exclusion and mandatory pre-read in your skill documentation includes a valid arrow pointer to a target skill-name, allowing agents to route correctly without hitting dead ends.

When should I split an overgrown agent skill instead of editing it?▼

You should split an overgrown agent skill when MECE enforcement and split detection checks reveal trigger overlap or scope violations against adjacent skills in the AGENTS.md index, indicating the skill has exceeded its boundaries.

What are the mandatory frontmatter and length requirements for agent skill directories?▼

Agent skill directories require frontmatter with name and description fields, mandatory routing pointers, scope partitioning, and a hard SKILL.md length cap to keep the skill scannable, testable, and correctly scoped.

What is the best way to validate a new skill against existing agent workflows?▼

The best way to validate a new skill against existing agent workflows is to check the AGENTS.md skill index and adjacent scope contracts, ensuring your skill has no trigger overlap, gaps, or routing dead ends.

Related Skills

Explore More Software Engineering Skills→
saint2706saint2706

writing-skills

Guide agent skill creation and testing with Test-Driven Development principles.

Community
Advanced
nibojniboj
1

skill-definir-perimetre

Define a scoped skill concept with trigger conditions and non-goals.

Community
Intermediate
xtrm-devxtrm-dev
4

agent-docs-maintainer

Audit and rewrite agent-facing project documentation into compact routing-oriented guides.

Official
Advanced
LISANG3LISANG3
1

writing-skills

Apply TDD principles to create and maintain agent Skill documentation.

Community
Advanced