benchmark-patterns

What problem does it solve?

Technical decisions are often based on opinions rather than measurable data, leading to suboptimal choices and costly rework. This skill provides a methodology for robust comparative benchmarks.

Core Features & Use Cases

• Multi-Dimensional Metrics: Defines metrics for performance (throughput, latency), developer experience (DX), ecosystem (GitHub stars, community), cost, and team fit.
• Standardized Methodology: Provides tools and methods for running benchmarks (e.g., wrk, k6, sysbench).
• Evaluation Matrix: Guides on scoring and weighting criteria to calculate a total score for each solution.
• Use Case: When evaluating two backend frameworks, use this skill to define benchmark criteria, run performance tests with wrk, assess developer experience, and calculate a weighted score, leading to a data-driven architectural decision documented in an ADR.

Quick Start

Use the benchmark-patterns skill to understand the key metrics for evaluating the performance of a technical solution.