Suggest Optimizations

Analyze code and profile data to recommend optimization strategies for improving performance and resource usage.

When to Use

• Performance-critical code paths identified in profiling
• Reducing execution time for hot functions
• Lowering memory usage
• Planning SIMD or vectorization strategies

Quick Reference

# Identify optimization opportunities
# 1. Profile to find bottlenecks
# 2. Analyze algorithmic complexity
# 3. Check for unnecessary operations
# 4. Evaluate data structure choices
# 5. Consider SIMD/vectorization

# Profile Python code
python3 -m cProfile -s cumulative script.py | head -20

Workflow

1. Profile critical paths: Identify functions consuming most time/memory
2. Analyze algorithms: Check time/space complexity, look for inefficiencies
3. Examine data structures: Verify optimal data structure choices
4. Consider caching: Identify repeated computations
5. Propose optimizations: List specific changes with expected impact

Output Format

Optimization recommendation:

• Bottleneck identified (function, line number)
• Current performance (time/memory)
• Root cause analysis
• Recommended optimization technique
• Expected improvement (percentage or time estimate)
• Implementation difficulty (low/medium/high)

References

• See benchmark-functions skill for measuring improvements
• See profile-code skill for detailed profiling
• See CLAUDE.md > Mojo for SIMD optimization patterns