Enhance Component

Analyze a Claude Code component's quality and get prioritized improvement suggestions.

Arguments:

• $1 (required): Component type - agent, skill, command, hook
• $2 (required): Component name to analyze

Full arguments: $ARGUMENTS

Workflow

1. Locate Component

   • Find component by type and name
   • Read current configuration

2. Delegate Analysis

   • Route via Task to appropriate builder
   • Builder performs deep quality analysis

3. Return Recommendations

   • Detailed scoring across dimensions
   • Prioritized improvement suggestions
   • Actionable next steps

Scoring Dimensions

Agents (40 points total)

Dimension	Points	What's Measured
Schema	10	Required fields, syntax
Security	10	Tool permissions, validation
Quality	10	Content, examples, workflow
Maintainability	10	Organization, documentation

Skills (40 points total)

Dimension	Points	What's Measured
Schema	10	No model field, structure
Auto-invoke	10	Trigger specificity
Resources	10	{baseDir} usage, organization
Documentation	10	Completeness, examples

Commands (40 points total)

Dimension	Points	What's Measured
Schema	10	Model format, fields
Arguments	10	Documentation, handling
Security	10	Bash validation
Usability	10	Examples, clarity

Hooks (40 points total)

Dimension	Points	What's Measured
Schema	10	JSON validity
Security	10	7-category analysis
Matchers	10	Specificity, safety
Scripts	10	Existence, permissions

Examples

Enhance an Agent

/agent-builder:enhance agent code-reviewer

Enhance a Skill

/agent-builder:enhance skill analyzing-code

Enhance a Command

/agent-builder:enhance command run-tests

Enhance a Hook

/agent-builder:enhance hook validate-write

Output Format

## Enhancement Analysis: [component-name]

**Overall Score**: X/40 (Y%)

### Detailed Scores
| Dimension | Score | Status |
|-----------|-------|--------|
| Schema | 10/10 | ✅ |
| Security | 8/10 | ⚠️ |
| Quality | 7/10 | ⚠️ |
| Maintainability | 9/10 | ✅ |

### Findings

#### ✅ Strengths
- [What's working well]

#### ⚠️ Areas for Improvement
- [What needs attention]

### Recommendations (Prioritized)

1. 🔴 HIGH: [Critical improvement]
2. 🟡 MEDIUM: [Important improvement]
3. 🟢 LOW: [Nice-to-have improvement]

### Next Steps
1. [Actionable step 1]
2. [Actionable step 2]

Execution

When invoked:

1. Parse type and name from arguments
2. Delegate via Task to appropriate builder
3. Builder performs comprehensive analysis
4. Return detailed findings and recommendations