← ClaudeAtlas

writing-skillslisted

Use when creating new skills, editing existing skills, or verifying skills work before deployment
gorban-dev/gor-mobile · ★ 2 · AI & Automation · score 75
Install: claude install-skill gorban-dev/gor-mobile
# Writing Skills ## Overview **Writing skills is baseline-first: observe what agents do WITHOUT the skill, then write the skill to close that gap.** **Personal skills live in agent-specific directories (`~/.claude/skills` for Claude Code, `~/.agents/skills/` for Codex)** You run pressure scenarios with subagents, watch them fail (baseline behavior), write the skill (documentation), watch agents comply, and harden (close loopholes). **Core principle:** If you didn't watch an agent fail without the skill, you don't know if the skill teaches the right thing. **Official guidance:** For Anthropic's official skill authoring best practices, see anthropic-best-practices.md. This document provides additional patterns and guidelines that complement the baseline-first approach in this skill. ## What is a Skill? A **skill** is a reference guide for proven techniques, patterns, or tools. Skills help future Claude instances find and apply effective approaches. **Skills are:** Reusable techniques, patterns, tools, reference guides **Skills are NOT:** Narratives about how you solved a problem once ## The Baseline-First Method | Step | Skill Creation | |------|----------------| | **Pressure scenario** | Subagent task that tempts the failure | | **The artifact** | Skill document (SKILL.md) | | **Baseline** | Agent violates the rule without the skill — document it | | **Verified** | Agent complies with the skill present | | **Harden** | Close loopholes while keeping compliance | |