sglang-prod-incident-triage
Installation
SKILL.md
SGLang Serving Debug
Overview
Use this skill to turn a live serving problem into a debug path you can replay.
Use one loop:
- collect a baseline bundle
- save the failing request or crash dump
- replay on a clean target
- only then switch tools
Do not start with profiling.
As of SGLang v0.5.17 / v0.5.18, two extra live-server signals matter: