skills.rest
dougborg
dougborgCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

skill-writer

Draft SKILL.md and agent files with progressive disclosure sections.

Updated Jul 18, 2025
Search Tags:#documentation#skill authoring#agent design#prompt structure#progressive disclosure#tool permissions#safety constraints
One-click install
npx skills add https://github.com/dougborg/stocktrim-openapi-client --skill skill-writer-dougborg
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-writer
Source: https://github.com/dougborg/stocktrim-openapi-client/tree/main/.claude/skills/skill-writer
Command: npx skills add https://github.com/dougborg/stocktrim-openapi-client --skill skill-writer-dougborg

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The skill creates consistent, progressive-disclosure skill and agent docs so humans and AI can quickly determine relevance, constraints, and next steps without reading long instructions.

Core Features & Use Cases

  • Progressive disclosure structure: Produces PURPOSE, CRITICAL, ASSUMES, STANDARD PATH, EDGE CASES, and optional DETAIL sections for fast skimming.
  • Safety-first drafting: Places catastrophic constraints in CRITICAL and separates responsibilities to prevent overly broad or unsafe behaviors.
  • Tool permission guidance: Sets allowed-tools to minimum necessary permissions and keeps execution boundaries clear.

Quick Start

Use the skill-writer to generate a new .claude/skills/<new-skill>/SKILL.md and matching agent file that follows PURPOSE, CRITICAL, ASSUMES, STANDARD PATH, EDGE CASES, and DETAIL conventions.

Frequently Asked Questions about skill-writer

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

FAQPage Schema
How do I structure agent documentation for progressive disclosure and token-budgeted sessions?▼

Agent documentation for progressive disclosure uses PURPOSE, CRITICAL, ASSUMES, STANDARD PATH, EDGE CASES, and optional DETAIL sections. This structure allows humans and AI to quickly determine relevance and constraints without reading long instructions during token-budgeted sessions.

What is the best way to enforce safety constraints and minimum tool permissions in skill authoring?▼

Safety constraints in skill authoring are enforced by placing catastrophic limitations in a CRITICAL section and setting allowed-tools to minimum necessary permissions. This separates responsibilities to prevent overly broad or unsafe agent behaviors.

How do I generate scannable SKILL.md files that define clear workflows and boundaries?▼

Scannable SKILL.md files are generated using a standardized progressive disclosure structure that defines constraints, assumptions, and bounded workflows. This format ensures fast skimming and enforces safety guardrails for both human and AI discovery.

When do I need to use progressive disclosure structure for skill and agent design?▼

Progressive disclosure structure is needed when creating new .claude/skills/*/SKILL.md and .claude/agents/*.md files. It standardizes documentation so humans and AI can quickly determine relevance, constraints, and next steps without reading long instructions.

Does skill authoring with bounded workflow length limit edge case documentation?▼

Bounded workflow length keeps standard paths concise but does not limit edge case documentation. Optional DETAIL sections are used specifically for referenced edge cases, ensuring the main workflow remains scannable while still covering necessary complex scenarios.

Related Skills

Explore More Software Engineering Skills→
duc01226duc01226

create-skill

Create Claude Skill scaffolds conforming to the Agent Skills open standard with required frontmatter fields and directory structure guidance.

Community
Intermediate
johanruttensjohanruttens

skill-writer

Create and validate Claude Code Agent Skills with SKILL.md frontmatter.

Community
Basic
Blake-JohnBlake-John

prompt-engineer

Write, refactor, and evaluate LLM prompts with JSON schemas.

Community
Advanced
Entropy1110Entropy1110
3

Mali Report Generation

Generate markdown vulnerability analysis reports from session outputs.

Community
Basic