Python & Rust Performance Hints (Jeff Dean & Sanjay Ghemawat style)

This skill packages key ideas from Abseil's Performance Hints document, adapted for Python and Rust development.

Use it to:

• review Python/Rust code for performance risks
• propose high-impact optimizations with explicit tradeoffs
• design APIs/data structures that keep future optimizations possible
• write an experiment plan (profile + microbenchmark) to validate changes

Scope and guardrails

• Scope: single-process / single-binary performance (CPU, memory, allocations, cache behavior).
• Do not: change externally observable behavior unless the user explicitly agrees.
• Do not: introduce undefined behavior, data races, or brittle "clever" micro-opts without evidence.
• Default philosophy: choose the faster alternative when it doesn't materially harm readability or complexity; otherwise, measure first.

When to apply