1. Define precise triggers, non-triggers, outcome, and required capabilities. 2. Keep the SKILL.md focused. Move heavy examples, references, and scripts into progressive-disclosure files. 3. Use portable capability language. Add a fallback for optional capabilities instead of hard-coding one host tool. 4. Validate Agent Skills frontmatter and directory/name parity. 5. Exercise the skill on representative positive and negative cases. Revise based on behavior, not aesthetics alone.
Installs into .claude/skills of the current project.
Are you the author of Playbooks?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/theonlyfusioncube-playbooks)
# Authoring a skill
1. Define precise triggers, non-triggers, outcome, and required capabilities.
2. Keep the SKILL.md focused. Move heavy examples, references, and scripts into progressive-disclosure files.
3. Use portable capability language. Add a fallback for optional capabilities instead of hard-coding one host tool.
4. Validate Agent Skills frontmatter and directory/name parity.
5. Exercise the skill on representative positive and negative cases. Revise based on behavior, not aesthetics alone.
