api-performance

Diagnose and optimize API latency and throughput for backend services.

Updated Mar 29, 2026

Search Tags:#caching #profiling #tracing #latency #throughput #api-performance #http2

Ready to run

npx skills add https://github.com/marquesfelip/agents-and-skills --skill api-performance

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: api-performance
Source: https://github.com/marquesfelip/agents-and-skills/tree/main/skills/api-performance
Command: npx skills add https://github.com/marquesfelip/agents-and-skills --skill api-performance

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

API latency optimization and throughput strategies for backend services.

Core Features & Use Cases

Profile bottlenecks with profiling and tracing to identify top offenders
Parallelize independent downstream calls using errgroup and bounded concurrency
Optimize payloads with sparse fieldsets, efficient serialization, and compression
Implement HTTP caching with ETag and Cache-Control to reduce repeated work
Apply timeouts and circuit breakers to bound cascading failures
Instrument and monitor performance with distributed tracing and dashboards

Quick Start

Run a targeted API performance optimization plan for a given slow endpoint and implement parallelism, caching, and timeout strategies.

Frequently Asked Questions about api-performance

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I reduce API latency for backend services with multiple downstream dependencies?▼

Reduce API latency by profiling bottlenecks, parallelizing independent downstream calls with errgroup, enforcing per-request deadlines, and applying robust timeouts to prevent cascading failures.

What is the best way to optimize API throughput and prevent cascading failures?▼

Optimize API throughput by bounding cascading failures with circuit breakers, applying HTTP caching using ETag and Cache-Control, and optimizing payloads with sparse fieldsets and compression.

How does distributed tracing work to identify API performance bottlenecks?▼

Distributed tracing profiles top offenders across backend services by instrumenting requests, allowing you to monitor performance and pinpoint slow downstream dependencies within your architecture.

Can I parallelize downstream API calls to improve backend performance?▼

Yes, you can parallelize independent downstream API calls using errgroup and bounded concurrency to execute requests simultaneously while enforcing per-request deadlines.

When should I apply circuit breakers and timeouts for API performance tuning?▼

Apply circuit breakers and robust timeouts when backend APIs have multiple downstream dependencies to bound cascading failures, enforce per-request deadlines, and maintain throughput under load.