Skip to main content

sglang

Fast structured generation and serving for LLMs with RadixAttention prefix caching. Use for JSON/regex outputs, constrained decoding, agentic workflows with too

89/100

Security score

The sglang skill was audited on Feb 28, 2026 and we found 7 security issues across 2 threat categories. Review the findings below before installing.

Categories Tested

Security Issues

medium line 384

Curl to non-GitHub URL

SourceSKILL.md
384curl http://localhost:30000/v1/chat/completions \
low line 43

External URL reference

SourceSKILL.md
43pip install sglang[all] flashinfer -i https://flashinfer.ai/whl/cu121/torch2.4/
low line 78

External URL reference

SourceSKILL.md
78sgl.set_default_backend(sgl.OpenAI("http://localhost:30000/v1"))
low line 384

External URL reference

SourceSKILL.md
384curl http://localhost:30000/v1/chat/completions \
low line 398

External URL reference

SourceSKILL.md
398client = OpenAI(base_url="http://localhost:30000/v1", api_key="EMPTY")
low line 438

External URL reference

SourceSKILL.md
438- **Docs**: https://sgl-project.github.io/
low line 440

External URL reference

SourceSKILL.md
440- **Discord**: https://discord.gg/sglang
Scanned on Feb 28, 2026
View Security Dashboard