Learn-skills.dev skill-creator
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.
git clone https://github.com/NeverSight/learn-skills.dev
T=$(mktemp -d) && git clone --depth=1 https://github.com/NeverSight/learn-skills.dev "$T" && mkdir -p ~/.claude/skills && cp -r "$T/data/skills-md/404kidwiz/claude-supercode-skills/skill-creator" ~/.claude/skills/neversight-learn-skills-dev-skill-creator-0e3a23 && rm -rf "$T"
data/skills-md/404kidwiz/claude-supercode-skills/skill-creator/SKILL.mdSkill Creator
Purpose
A meta-skill that guides the creation of high-quality, effective skills. Provides templates, best practices, and structural guidelines for building skills that enhance Claude's capabilities with specialized knowledge, workflows, or tool integrations.
When to Use
- User wants to create a new skill
- User wants to update or improve an existing skill
- User asks how to structure skill documentation
- Need to design a skill for a specific domain or workflow
- Want to ensure skill follows best practices
Core Skill Structure
Required Components
Every skill must have these elements:
-
Frontmatter
--- name: skill-name description: One-line description when to use this skill --- -
Title & Purpose
# Skill Name ## Purpose Clear, concise statement of what this skill does -
When to Use
## When to Use - Specific trigger 1 - Specific trigger 2 - Context where this helps -
Core Capabilities
## Core Capabilities ### Domain Expertise - Key knowledge area 1 - Key knowledge area 2 ### Tools & Methods - Specific techniques - Frameworks used
Optional but Recommended Components
-
Workflow
## Workflow 1. Step 1: What to do first 2. Step 2: Next action 3. Step 3: Final deliverable -
Best Practices
## Best Practices - Do this - Avoid that - Remember this -
Examples
## Examples ### Example 1: Common Use Case **Input**: User request **Approach**: How to handle **Output**: Expected result -
Anti-Patterns
## Anti-Patterns ❌ **Don't**: Bad practice ✅ **Do**: Good alternative
Skill Creation Workflow
Step 1: Define Scope
Ask yourself:
- What problem does this skill solve?
- Who will use it?
- What triggers its use?
- What's the expected outcome?
Step 2: Identify Core Knowledge
Document:
- Domain-specific terminology
- Key concepts and principles
- Common patterns in this domain
- Tools and technologies involved
Step 3: Structure the Workflow
Map out:
- Entry conditions
- Step-by-step process
- Decision points
- Exit criteria and deliverables
Step 4: Add Practical Elements
Include:
- Real-world examples
- Common pitfalls to avoid
- Best practices from the field
- Quality criteria
Step 5: Write Clear Triggers
Make "When to Use" specific:
-
✅ "User needs SQL query optimization for PostgreSQL databases"
-
❌ "User needs database help"
-
✅ "Debugging production outages in distributed systems"
-
❌ "Fixing bugs"
Skill Quality Criteria
Clarity
- Name is self-explanatory
- Description clearly states when to use
- Purpose is stated in 1-2 sentences
- No jargon without explanation
Completeness
- All required sections present
- Workflow is actionable
- Examples cover common cases
- Edge cases addressed
Specificity
- Triggers are concrete
- Steps are detailed enough to follow
- Tools/methods are named explicitly
- Success criteria defined
Usability
- Easy to scan and navigate
- Consistent formatting
- Logical section ordering
- Cross-references where helpful
Skill Templates
Technical Domain Skill Template
--- name: domain-expert description: Use when user needs [specific technical task] in [technology/domain] --- # Domain Expert ## Purpose Expert in [domain] specializing in [specific areas]. Helps with [key problems solved]. ## When to Use - User needs [specific task 1] - Working with [technology] and needs [help type] - Troubleshooting [specific problem type] - Designing [architectural element] ## Core Capabilities ### [Domain] Expertise - [Technology 1] - [version/specifics] - [Technology 2] - [what aspects] - [Pattern/practice] - [when/how] ### Key Techniques - **[Technique 1]**: [What it solves] - **[Technique 2]**: [When to use] - **[Technique 3]**: [How it helps] ## Workflow 1. **Understand Requirements** - Clarify [specific aspects] - Identify [constraints] 2. **Apply [Domain] Patterns** - Use [pattern 1] for [scenario] - Consider [trade-off] 3. **Implement Solution** - Follow [best practice] - Ensure [quality criteria] 4. **Validate** - Test [aspects] - Verify [requirements met] ## Best Practices - **[Practice 1]**: [Reasoning] - **[Practice 2]**: [Benefit] - **[Practice 3]**: [Why important] ## Common Patterns ### [Pattern 1] **When**: [Scenario] **How**: [Implementation approach] **Why**: [Benefits] ### [Pattern 2] **When**: [Scenario] **How**: [Implementation approach] **Why**: [Benefits] ## Anti-Patterns ❌ **Don't**: [Bad practice] - Why it fails: [Reason] - Better approach: [Alternative] ❌ **Avoid**: [Common mistake] - Problem: [What goes wrong] - Instead: [Correct way] ## Examples ### Example 1: [Common Scenario] **Context**: [Situation] **Approach**: [Solution steps] **Result**: [Outcome] ## Tools & Technologies - **[Tool 1]**: [Version] - [Use for what] - **[Tool 2]**: [Version] - [Use for what] - **[Framework]**: [Version] - [Key features used]
Process/Workflow Skill Template
--- name: process-specialist description: Use when user needs [specific process/workflow] for [outcome] --- # Process Specialist ## Purpose Guides [specific process] to achieve [specific outcome]. Ensures [quality aspects] through [methodology]. ## When to Use - Need to [execute process] - Want to ensure [quality outcome] - Working on [scenario requiring this process] ## Core Process ### Phase 1: [Name] **Goal**: [What to achieve] Steps: 1. [Action 1]: [Details] 2. [Action 2]: [Details] 3. [Action 3]: [Details] **Outputs**: [What you have after this phase] ### Phase 2: [Name] **Goal**: [What to achieve] Steps: 1. [Action 1]: [Details] 2. [Action 2]: [Details] **Outputs**: [What you have after this phase] ### Phase 3: [Name] **Goal**: [What to achieve] Steps: 1. [Action 1]: [Details] 2. [Action 2]: [Details] **Deliverable**: [Final output] ## Decision Points ### When to [Decision] - If [condition], then [choice A] - If [condition], then [choice B] ## Quality Gates After each phase, verify: - [ ] [Criterion 1] - [ ] [Criterion 2] - [ ] [Criterion 3] ## Best Practices - **[Practice]**: [Why it matters] - **[Practice]**: [Impact on quality] ## Common Pitfalls - **Pitfall**: [What people do wrong] - **Impact**: [What happens] - **Solution**: [How to avoid]
Writing Tips
Be Specific
❌ "Use when working with databases" ✅ "Use when optimizing SQL queries for PostgreSQL 14+ production databases"
Be Actionable
❌ "Think about security" ✅ "Run OWASP ZAP scan and review all HIGH severity findings"
Be Structured
Use consistent heading levels:
for major sections##
for subsections###
for detailed breakdowns####
Use Visual Indicators
- ✅ for good practices
- ❌ for anti-patterns
- 🔍 for investigation steps
- ⚠️ for warnings
- 💡 for tips
Include Context
Don't just list what to do—explain why:
## Instead of: - Use connection pooling ## Write: - **Use connection pooling** (pg-pool for PostgreSQL) - Reduces connection overhead by 80% - Critical for applications with >100 concurrent users - Configure pool size = (core count × 2) + effective_spindle_count
Skill Maintenance
When to Update
- New version of core technology released
- Better practices emerge in the field
- User feedback reveals gaps
- Related skills are created (cross-reference)
Version Control
Consider adding to frontmatter:
--- name: skill-name description: One-line description ---
Skill Integration
Cross-References
Link to related skills:
## Related Skills - Use [[debugger-skill]] when issues arise - Combine with [[performance-engineer-skill]] for optimization - Precede with [[architect-reviewer-skill]] for design validation
Skill Composition
Complex workflows can chain skills:
## Workflow 1. Use [[requirement-analyst]] to gather needs 2. Apply this skill for implementation 3. Use [[code-reviewer]] for quality assurance 4. Use [[deployment-engineer]] to ship
Examples
Example 1: Creating a Python Pro Skill
Context: Need a skill for advanced Python development
Process:
- Define scope: Python 3.11+ with focus on FastAPI and type safety
- Identify triggers: "modern Python", "type hints", "FastAPI"
- Structure core capabilities:
- Python 3.11+ features (match statements, typing improvements)
- FastAPI framework patterns
- Type annotation best practices
- Add workflow: Design API → Type models → Implement routes → Test
- Include examples: FastAPI route with full type annotations
Result: A focused, actionable skill for modern Python development
Example 2: Creating a Git Workflow Skill
Context: Need to codify team's git branching strategy
Process:
- Define scope: Git workflow for feature development
- Identify triggers: "create branch", "make PR", "git workflow"
- Structure as phases:
- Branch creation
- Development cycle
- PR process
- Merge strategy
- Add decision points: When to rebase vs merge
- Include examples: Standard feature development flow
Result: Clear procedural guide for consistent git usage
Validation Checklist
Before finalizing a skill, check:
Structure
- Frontmatter complete (name, description)
- Title and purpose clear
- "When to Use" section has specific triggers
- Core capabilities well-defined
Content
- Information is accurate and current
- Examples are realistic and helpful
- Best practices are justified
- Anti-patterns show alternatives
Usability
- Can scan and find info quickly
- Sections flow logically
- Formatting is consistent
- Cross-references are correct
Quality
- No spelling/grammar errors
- Technical terms defined
- Code examples (if any) are correct
- Meets all quality criteria above
Meta: About This Skill
This skill itself demonstrates the principles it teaches:
- Clear frontmatter and structure
- Specific "When to Use" triggers
- Actionable workflows
- Concrete examples
- Quality criteria
When creating skills, use this as both a guide and a template.