generate-profile

Profile SGLang server runs and capture Chrome-compatible traces for performance analysis.

1|Updated May 8, 2026
One-click install
npx skills add https://github.com/dyyoungg/sglang-dev --skill generate-profile-dyyoungg
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: generate-profile
Source: https://github.com/dyyoungg/sglang-dev/tree/main/.claude/skills/generate-profile
Command: npx skills add https://github.com/dyyoungg/sglang-dev --skill generate-profile-dyyoungg

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill provides an end-to-end profiling workflow for SGLang server runs, enabling developers to identify perf bottlenecks and latency sources by generating Chrome-compatible traces.

Core Features & Use Cases

  • End-to-end profiling: Launch a server, validate readiness, and capture a Chrome-compatible trace for performance analysis.
  • Guided workflow: Step-by-step instructions to reproduce profiling in local or GPU-enabled environments.
  • Use Case: Use profiling to optimize startup time, throughput, and memory usage in model-serving deployments.

Quick Start

Run the profiling workflow against a live SGLang server to produce a trace.

Frequently Asked Questions about generate-profile

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

FAQPage Schema
How do I profile an SGLang server to identify performance bottlenecks?

SGLang server profiling captures Chrome-compatible traces end-to-end to identify performance bottlenecks and latency sources. It orchestrates server launch, health validation, trace collection, and reporting through a reproducible workflow for both local and GPU-enabled deployments.

What is the best way to capture a trace for SGLang performance analysis?

Capturing a trace for SGLang performance analysis involves running an end-to-end profiling workflow that launches the server, validates readiness, and collects Chrome-compatible traces. This guided process ensures reproducible profiling results for optimizing startup time, throughput, and memory usage.

Can I use this profiling workflow in a local development environment?

Yes, the SGLang profiling workflow applies to local development and GPU-enabled deployments. It provides step-by-step instructions to reproduce server profiling and capture Chrome-compatible traces regardless of your environment setup.

Why do I need a Chrome-compatible trace for my model serving deployment?

A Chrome-compatible trace is needed to visualize and analyze performance bottlenecks in your model-serving deployment. It pinpoints latency sources affecting SGLang server startup time, throughput, and memory usage.

Does SGLang profiling optimize startup time and memory usage automatically?

SGLang profiling does not optimize automatically; it identifies performance bottlenecks by generating a trace. You analyze the trace results to manually optimize SGLang server startup time, throughput, and memory usage.