skills.rest
LisztVan
LisztVanCOMMUNITY
Explore all skills by this author →
GitHub

writing-skills

Create and validate reusable Skill documentation for agent workflows.

2|Updated Apr 24, 2026
Search Tags:#tdd#documentation#prompt engineering#verification#subagents#skill authoring
One-click install
npx skills add https://github.com/LisztVan/subcli-cpp --skill writing-skills-lisztvan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: writing-skills
Source: https://github.com/LisztVan/subcli-cpp/tree/main/.pi/skills/superpowers/skills/writing-skills
Command: npx skills add https://github.com/LisztVan/subcli-cpp --skill writing-skills-lisztvan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps create, edit, and verify Skill documentation so agents can reliably discover the right skill, follow it under pressure, and avoid loopholes that weaken compliance.

Core Features & Use Cases

  • Skill authoring guidance for new skills, including naming, discovery-friendly descriptions, and compact frontmatter.
  • Validation and testing discipline for checking skills with baseline scenarios, pressure tests, and refinement cycles before deployment.
  • Documentation architecture for keeping core instructions inline while splitting heavy references, reusable tools, and flowcharts into the right places.
  • Use case: A team building agent workflows uses this skill to turn informal know-how into a reusable instruction set that other agents can find and apply consistently.

Quick Start

Use this skill to draft or revise a SKILL.md file, then test it with realistic scenarios before publishing it.

Frequently Asked Questions about writing-skills

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

FAQPage Schema
How do I write agent instructions that LLMs actually follow consistently?▼

To write agent instructions that LLMs follow consistently, draft a SKILL.md file using progressive disclosure, discovery-friendly metadata, and loophole-resistant instructions to enforce strict compliance.

What is TDD-style skill validation for prompt engineering?▼

TDD-style skill validation is a red-green-refactor discipline that pressure-tests agent instructions against baseline scenarios with subagents before deployment to identify and refine compliance loopholes.

How do you structure documentation architecture for an agent workflow?▼

Documentation architecture for an agent workflow keeps core instructions inline while splitting heavy references, reusable tools, and flowcharts into separate files to maintain agent context efficiency.

What's the best way to test agent prompts before deployment?▼

The best way to test agent prompts before deployment is running pressure tests with subagents using a TDD-style discipline, applying baseline scenarios to evaluate instruction compliance and refine loopholes.

Why do agents ignore instructions in my prompt engineering workflow?▼

Agents ignore instructions due to loophole-resistant gaps, poor discovery-friendly metadata, or overloaded context; fixing this requires drafting compact frontmatter and applying progressive disclosure architecture.

Related Skills

Explore More Software Engineering Skills→
dhar174dhar174
1

notion-knowledge-capture

Convert informal chats and notes into structured Notion pages.

Community
Intermediate
seanGSISGseanGSISG
3

testing-skills-with-subagents

Test skills with subagents to verify resistance to rationalization under pressure.

Community
Basic
robertcoperobertcope
1

Prompt Mastery

Craft structured Claude prompts using XML, chain-of-thought, and prefilling techniques.

Community
Advanced
samjhechtsamjhecht
4

analyzing-research-documents

Extract decisions, constraints, and action items from research documents.

Community
Advanced