Improve Component Quality

You are improving the quality of a Claude Code component by applying systematic enhancements based on quality analysis.

Task

Improve the component at path: $1

Process

1. Analyze Current Quality

First, analyze the component to identify improvement opportunities:

/analyze-component $1

Or directly invoke the analyzing-component-quality skill to get quality scores and improvement suggestions.

2. Plan Improvements

Based on the quality analysis, create an improvement plan:

## Improvement Plan for [Component Name]

**Current Quality**: X.X/5
**Target Quality**: X.X/5

### Priority 1: Critical Issues
- [ ] [Issue 1]
- [ ] [Issue 2]

### Priority 2: Important Issues
- [ ] [Issue 3]
- [ ] [Issue 4]

### Priority 3: Nice to Have
- [ ] [Issue 5]

3. Apply Improvements

Invoke the improving-components skill and apply improvements systematically:

A. Description Enhancement

If description is vague or missing triggers:

# Before
description: Helps with testing

# After (use Edit tool)
description: Expert at writing Jest unit tests for JavaScript/TypeScript. Auto-invokes when user writes new functions or asks "test this code". Generates comprehensive test suites with mocks, assertions, and edge cases following AAA pattern.

Apply using Edit tool:

Edit $1:
  old_string: "description: [old vague description]"
  new_string: "description: [specific detailed description]"

B. Tool Permission Optimization

If tools are excessive or unjustified:

# Before
allowed-tools: Read, Write, Edit, Bash, Grep, Glob, Task

# After (for research skill)
allowed-tools: Read, Grep, Glob, WebSearch, WebFetch

Apply using Edit tool:

Edit $1:
  old_string: "allowed-tools: [old excessive list]"
  new_string: "allowed-tools: [minimal necessary list]"

C. Documentation Enhancement

If missing examples or structure:

Add sections like:

## Capabilities

- [Specific capability 1]
- [Specific capability 2]

## Examples

### Example 1: [Scenario]

User: "How does X work?" Skill: [Action] Result: [Outcome]


### Example 2: [Scenario]
[Another example]

Apply using Edit tool to insert at appropriate location.

D. Security Improvements

If security concerns exist:

Remove unnecessary Bash access
Add input validation notes
Eliminate dangerous tool combinations
Add security documentation

4. Validate Improvements

After applying changes:

# Re-run quality analysis
python self-improvement/skills/analyzing-component-quality/scripts/quality-scorer.py "$1"

Verify:

✓ Quality score improved
✓ Critical issues resolved
✓ No new issues introduced
✓ Component still functional

5. Present Results

Show a comprehensive before/after report:

# Component Improvement Report

**Component**: [name]
**Type**: [type]
**Location**: `$1`

## Changes Applied

### 1. Description Enhanced
**Before**:
```yaml
description: Helps with code

After:

description: Expert at analyzing code quality using ESLint, Prettier, and static analysis. Auto-invokes when user finishes writing code or asks "is this code good?". Provides actionable improvement suggestions with severity levels.

Impact: Description Clarity improved from 2/5 to 5/5 (+3 points)

2. Tool Permissions Optimized

Before: Read, Write, Edit, Bash, Grep, Glob, Task After: Read, Bash

Removed:

Write, Edit: Code analysis doesn't modify files
Grep, Glob: Linters handle searching
Task: Skills don't delegate to agents

Kept:

Read: Needed to read code files
Bash: Needed to run linters (eslint, prettier)

Impact:

Tool Permissions improved from 2/5 to 4/5 (+2 points)
Security improved from 3/5 to 4/5 (+1 point)

3. Documentation Added

Added:

Capabilities section
3 usage examples
When to use vs. when not to use
Integration examples

Impact: Usability improved from 3/5 to 5/5 (+2 points)

Quality Score Improvement

Dimension	Before	After	Change
Description Clarity	2/5	5/5	+3
Tool Permissions	2/5	4/5	+2
Auto-Invoke Triggers	2/5	5/5	+3
Security	3/5	4/5	+1
Usability	3/5	5/5	+2
Overall	2.4/5	4.6/5	+2.2

Quality Level: Poor → Excellent ✅

Remaining Recommendations

[Any suggestions not automatically applied, if any]

Files Modified

$1 - Component improved

Next Steps

Review changes: Examine the diff to ensure all changes are appropriate
Test component: Verify it still works as intended

Commit changes:

git add $1
git commit -m "feat(self-improvement): enhance [component-name] quality

- Improved description clarity and specificity
- Optimized tool permissions for security
- Added comprehensive documentation and examples

Quality score: 2.4/5 → 4.6/5 (+2.2 points)"

Validate with agent-builder (optional):

python agent-builder/skills/building-[type]s/scripts/validate-[type].py "$1"


## Improvement Guidelines

### What to Improve

**Always improve**:
- Vague descriptions → Specific descriptions
- Excessive tools → Minimal necessary tools
- Security vulnerabilities → Secure patterns
- Missing documentation → Comprehensive docs

**Carefully improve**:
- Existing functionality (preserve behavior)
- Working auto-invoke triggers (ensure backward compatible)
- Tool lists (verify each change necessary)

**Never change**:
- Component name (breaking change)
- Core functionality (preserve intent)
- Version number (user manages this)

### Safety Checks

Before applying each improvement:
- [ ] Preserves component functionality
- [ ] Doesn't break existing usage
- [ ] Improves quality score
- [ ] No new issues introduced
- [ ] Changes are justified

### When to Ask User

Ask user confirmation before:
- Removing tools that might be needed
- Making changes that could affect existing usage
- Large-scale restructuring
- Uncertain about intent

**Example**:

I've identified that the Task tool may be unnecessary for this skill. Removing it would improve security score from 3/5 to 5/5.

Current: allowed-tools: Read, Grep, Glob, Task Proposed: allowed-tools: Read, Grep, Glob

Should I remove the Task tool? (y/n)


## Advanced Improvements

### Pattern-Based Improvements

Apply common improvement patterns:

**Pattern 1**: Research Skill Optimization

IF component_type == 'skill' AND 'research' in description: KEEP: Read, Grep, Glob, WebSearch, WebFetch REMOVE: Write, Edit, Bash, Task


**Pattern 2**: Code Generation Skill Optimization

IF component_type == 'skill' AND 'generate' in description: KEEP: Read, Write, Grep, Glob REMOVE: Bash (unless running tests), Task


**Pattern 3**: Agent Tool Optimization

IF component_type == 'agent': REMOVE: Task (agents shouldn't delegate to agents)


### Bulk Improvements

If multiple components need similar improvements:

```markdown
Detected pattern: 5 skills have Task tool but don't delegate to agents

Apply bulk improvement?
- research-skill: Remove Task tool
- analysis-skill: Remove Task tool
- pattern-skill: Remove Task tool
- best-practices-skill: Remove Task tool
- investigation-skill: Remove Task tool

Total impact: +10 security points across 5 components

Apply to all? (y/n)

Output Format

Provide a comprehensive improvement report that:

✓ Shows before/after for each change
✓ Quantifies quality improvements
✓ Explains rationale for changes
✓ Lists all modified files
✓ Provides next steps
✓ Includes validation results

Important Notes

Preserve functionality: Never change what the component does
Incremental changes: Apply one category at a time
Validate after: Always re-run quality scoring
Show evidence: Demonstrate quality improved
Be conservative: When unsure, ask user
Document reasoning: Explain why each change helps

Your improvements make components more effective, secure, and maintainable for the entire Claude Code ecosystem.

/improve-component