skills.rest
Nabilhassan12345
Nabilhassan12345COMMUNITY
Explore all skills by this author →
GitHub

generate-profile

Launch an SGLang server and generate a Chrome profiling trace.

Updated Apr 7, 2026
Search Tags:#performance#profiling#trace#server#benchmark#cuda#sglang
One-click install
npx skills add https://github.com/Nabilhassan12345/voice-ai-workspace --skill generate-profile
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: generate-profile
Source: https://github.com/Nabilhassan12345/voice-ai-workspace/tree/main/sglang/.claude/skills/generate-profile
Command: npx skills add https://github.com/Nabilhassan12345/voice-ai-workspace --skill generate-profile

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill automates launching an SGLang server, validating its accuracy, generating a profiling trace, and returning the profile path.

Core Features & Use Cases

  • Launches an SGLang server with a configurable model and port.
  • Validates readiness and accuracy via a quick sanity check.
  • Generates a Chrome-compatible profiling trace and saves artifacts like the trace file and server_args.json.
  • Use Case: Diagnose performance characteristics of an SGLang deployment in local development or CI.

Quick Start

Launch the SGLang server with your model, wait for readiness, run the profiling, and retrieve the profile directory.

Frequently Asked Questions about generate-profile

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

FAQPage Schema
How do I generate a Chrome trace for an SGLang server?▼

Profiling an SGLang server automates launching the server, validating accuracy via a sanity check, and saving tracing data with server_args.json into a profile directory for benchmarking.

What is the best way to benchmark SGLang startup time and performance?▼

Benchmarking SGLang startup time is done by executing an end-to-end profiling workflow that produces a Chrome-compatible trace, capturing detailed execution data for performance analysis.

Do I need a CUDA GPU to profile an SGLang server?▼

Yes, profiling an SGLang server requires CUDA GPU availability and a working SGLang installation to execute the workflow and generate the Chrome trace successfully.

Can I run SGLang profiling across different models and ports?▼

Yes, you can profile SGLang across different models and ports by launching the server with configurable settings, making it suitable for local development or CI environments.

What artifacts are produced when tracing an SGLang server?▼

Tracing an SGLang server produces a profile directory containing a Chrome-compatible trace file and a server_args.json file, capturing profiling data and server configuration for benchmarking.

Related Skills

Explore More Software Engineering Skills→
QingHuan-6QingHuan-6

debug-cuda-crash

Diagnose CUDA crashes in SGLang by exposing kernel API call traces and tensor metadata.

Community
Intermediate
curiositechcuriositech
181

performance-profiling

Analyze Node.js application performance via CPU profiling and database query analysis.

Official
Advanced
bomberoxenviosdosruedasbomberoxenviosdosruedas

agency-devops-automator

Automate DevOps workflows with Terraform, GitHub Actions, Kubernetes, and Prometheus.

Community
Advanced
annealing-inversionannealing-inversion

generate-profile

Profile an SGLang server run and generate a Chrome trace and server_args.json.

Community
Advanced