
Create Agent Skills
If you're building Claude Code skills, this is the meta-skill you want. It walks you through creating new skills from scratch, auditing existing ones, and understanding the router pattern for complex skills. The intake is clean: it asks what you want to do, then routes you to specific workflows for creation, modification, or adding components like templates and scripts. What I appreciate is the progressive disclosure approach. It keeps SKILL.md under 500 lines and splits detailed content into references and workflows, so you're not drowning in documentation. The principles are solid too: pure XML structure, no markdown headings in the body, and treating skills as modular prompts. Worth using if you're authoring more than one or two skills.
- skills.sh installs
- 507
- GitHub stars
- 2K
- last updated
- Aug 21, 2026
Install
$ npx -y skills add glittercowboy/taches-cc-resources --skill create-agent-skills --agent claude-codeInstalls into .claude/skills of the current project.
Files
<essential_principles>
How Skills Work
Skills are modular, filesystem-based capabilities that provide domain expertise on demand. This skill teaches how to create effective skills.
1. Skills Are Prompts
All prompting best practices apply. Be clear, be direct, use XML structure. Assume Claude is smart - only add context Claude doesn't have.
2. SKILL.md Is Always Loaded
When a skill is invoked, Claude reads SKILL.md. Use this guarantee:
- Essential principles go in SKILL.md (can't be skipped)
- Workflow-specific content goes in workflows/
- Reusable knowledge goes in references/
3. Router Pattern for Complex Skills
skill-name/
├── SKILL.md # Router + principles
├── workflows/ # Step-by-step procedures (FOLLOW)
├── references/ # Domain knowledge (READ)
├── templates/ # Output structures (COPY + FILL)
└── scripts/ # Reusable code (EXECUTE)
SKILL.md asks "what do you want to do?" → routes to workflow → workflow specifies which references to read.
Sponsors
Details
- Repository
- github.com/glittercowboy/taches-cc-resources
- Categories
- Documentation, AI & Agent Building
- First seen
- Jun 3, 2026