---

name: performance-profiler description: CPU/memory profiling, database query optimization, and performance analysis version: 1.0.0 tags: [performance, profiling, optimization, cpu, memory, database]

Performance Profiler Skill

Purpose

The Performance Profiler Skill analyzes application performance, identifies bottlenecks, profiles CPU and memory usage, and optimizes database queries. It provides actionable insights to improve application speed and resource utilization.

Key Capabilities:

• CPU profiling and hotspot detection
• Memory profiling and leak detection
• Database query analysis and optimization
• Response time analysis
• Resource utilization monitoring
• Performance regression detection

Target Token Savings: 65% (from ~2200 tokens to ~770 tokens)

When to Use

• Investigating slow performance
• Optimizing application speed
• Detecting memory leaks
• Analyzing database queries
• Profiling API endpoints
• Monitoring resource usage
• Finding performance bottlenecks
• Regression testing

Operations

1. profile-cpu

Profiles CPU usage and identifies performance hotspots.

2. profile-memory

Analyzes memory usage and detects leaks.

3. analyze-queries

Examines database queries for optimization opportunities.

4. profile-api

Profiles API endpoint response times.

5. analyze-all

Comprehensive performance analysis.

Scripts

Copy# CPU profiling
python ~/.claude/skills/performance-profiler/scripts/main.py \
  --operation profile-cpu \
  --app-file app.py

# Memory profiling
python ~/.claude/skills/performance-profiler/scripts/main.py \
  --operation profile-memory \
  --app-file app.py

# Query analysis
python ~/.claude/skills/performance-profiler/scripts/main.py \
  --operation analyze-queries \
  --log-file queries.log

# Comprehensive analysis
python ~/.claude/skills/performance-profiler/scripts/main.py \
  --operation analyze-all \
  --app-file app.py

Configuration

Copy{
  "performance-profiler": {
    "cpu": {
      "threshold_percent": 80,
      "sample_interval": 0.01
    },
    "memory": {
      "threshold_mb": 100,
      "track_allocations": true
    },
    "queries": {
      "slow_query_threshold_ms": 100,
      "max_queries": 1000
    }
  }
}

Examples

Example 1: CPU Profiling

Copypython ~/.claude/skills/performance-profiler/scripts/main.py \
  --operation profile-cpu \
  --app-file app.py

Output:

Copy{
  "success": true,
  "operation": "profile-cpu",
  "total_time": 2.45,
  "hotspots": [
    {
      "function": "process_data",
      "time_percent": 45.2,
      "calls": 1000,
      "recommendation": "Consider caching or optimization"
    }
  ],
  "execution_time_ms": 234
}

Example 2: Memory Profiling

Copypython ~/.claude/skills/performance-profiler/scripts/main.py \
  --operation profile-memory \
  --app-file app.py

Output:

Copy{
  "success": true,
  "operation": "profile-memory",
  "peak_memory_mb": 156.7,
  "leaks_detected": 2,
  "recommendations": [
    "Large list allocation in loop - consider generator",
    "Unclosed database connections detected"
  ],
  "execution_time_ms": 456
}

Example 3: Query Analysis

Copypython ~/.claude/skills/performance-profiler/scripts/main.py \
  --operation analyze-queries \
  --log-file queries.log

Output:

Copy{
  "success": true,
  "operation": "analyze-queries",
  "total_queries": 458,
  "slow_queries": 12,
  "recommendations": [
    {
      "query": "SELECT * FROM users WHERE email = ...",
      "time_ms": 245,
      "issue": "Missing index on email column",
      "recommendation": "CREATE INDEX idx_users_email ON users(email)"
    }
  ],
  "execution_time_ms": 123
}

Example 4: Comprehensive Analysis

Copypython ~/.claude/skills/performance-profiler/scripts/main.py \
  --operation analyze-all \
  --app-file app.py

Output:

Copy{
  "success": true,
  "operation": "analyze-all",
  "summary": {
    "cpu_hotspots": 3,
    "memory_issues": 2,
    "slow_queries": 5,
    "overall_score": 72
  },
  "execution_time_ms": 1234
}

Token Economics

Without Skill: ~2200 tokens (manual analysis) With Skill: ~770 tokens (65% savings)

Success Metrics

• Execution time: <500ms for profiling
• Hotspot detection: >95% accuracy
• Memory leak detection: >90% accuracy
• Query optimization: 50-80% performance improvement

---

Performance Profiler Skill v1.0.0 - Optimizing application performance