← All skills
inference-serving-vllm
70 / 100 · Review before installing
Check before installing. Some of what it can do is expected: a coding skill usually runs build and test commands.
By davila7/claude-code-templates (GitHub) · coding · 33,256 stars · scanned at commit a6e6f93
Serves LLMs with high throughput using vLLM's PagedAttention and continuous batching. Use when deploying production LLM APIs, optimizing inference latency/throughput, or serving models with limited GPU memory. Supports OpenAI-compatible endpoints, quantization (GPTQ/AWQ/FP8), and tensor parallelism.
Works with any agent that supports SKILL.md skills.
What this skill can do
Downloads and installs other software
The author does not mention this.
Found in 8 places.
Show 5 more
Runs commands on your computer
The author does not mention this. Expected for a coding skill.
Found in 15 places.
Show 12 more
- cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 98
pip install locust - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 111
curl http://localhost:9090/metrics | grep vllm - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 125
docker run --gpus all -p 8000:8000 \ - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 251
vllm serve TheBloke/Llama-2-70B-AWQ \ - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 289
vllm serve MODEL \ - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 296
vllm serve MODEL --quantization awq - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 303
vllm serve MODEL --enable-prefix-caching - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 308
vllm serve MODEL --enable-chunked-prefill - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 315
vllm serve MODEL --trust-remote-code - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 322
vllm serve MODEL --max-num-seqs 512 - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 331
vllm serve MODEL --tensor-parallel-size 4 # Not 3 - cli-tool/components/skills/ai-research/inference-serving-vllm/SKILL.md, line 336
vllm serve MODEL --speculative-model DRAFT_MODEL
Can call paid AI services billed to you
This is disclosed by the author in the skill's SKILL.md or README.
Found in 3 places.
Sends data to these websites: api.openai.com
The author does not mention this.
Found in 14 places.
Show 11 more
- cli-tool/components/skills/ai-research/inference-serving-vllm/references/quantization.md, line 58
from openai import OpenAI - cli-tool/components/skills/ai-research/inference-serving-vllm/references/server-deployment.md, line 225
curl http://localhost:8000/health - cli-tool/components/skills/ai-research/inference-serving-vllm/references/server-deployment.md, line 232
until curl -f http://localhost:8000/health; do - cli-tool/components/skills/ai-research/inference-serving-vllm/references/troubleshooting.md, line 207
curl http://localhost:8000/health - cli-tool/components/skills/ai-research/inference-serving-vllm/references/troubleshooting.md, line 231
from openai import OpenAI - cli-tool/components/skills/ai-research/inference-serving-vllm/references/troubleshooting.md, line 251
session = requests.Session() - cli-tool/components/skills/ai-research/inference-serving-vllm/references/troubleshooting.md, line 377
curl http://localhost:9090/metrics - cli-tool/components/skills/ai-research/inference-serving-vllm/references/troubleshooting.md, line 380
curl http://localhost:9090/metrics | grep vllm_time_to_first_token - cli-tool/components/skills/ai-research/inference-serving-vllm/references/troubleshooting.md, line 394
curl http://localhost:8000/health - cli-tool/components/skills/ai-research/inference-serving-vllm/references/troubleshooting.md, line 397
curl http://localhost:8000/v1/models - cli-tool/components/skills/ai-research/inference-serving-vllm/references/troubleshooting.md, line 400
curl http://localhost:8000/v1/completions \
Safer alternatives
Low-risk skills in the same category (coding) that look related. Check that one does what you need.
Install
Paste this into your terminal. It downloads exactly the version we scanned, not whatever is newest.
if [ -e ~/.agents/skills/'inference-serving-vllm' ] || [ -L ~/.agents/skills/'inference-serving-vllm' ]; then echo 'A skill folder named inference-serving-vllm already exists — not replacing it' >&2; false; else d=$(mktemp -d) && git clone --quiet --filter=blob:none --no-checkout 'https://github.com/davila7/claude-code-templates' "$d" && git -C "$d" checkout --quiet 'a6e6f93e9a4baaa6ff337aee5505bf5c2bae4ce6' -- 'cli-tool/components/skills/ai-research/inference-serving-vllm' && mkdir -p ~/.agents/skills && cp -R "$d"/'cli-tool/components/skills/ai-research/inference-serving-vllm' ~/.agents/skills/'inference-serving-vllm' && rm -rf "$d"; fi
Codex reads ~/.agents/skills directly. For Claude Code, Cursor, Copilot, Gemini CLI, Antigravity or OpenCode, also link it. For Claude Code: ln -s ~/.agents/skills/'inference-serving-vllm' ~/.claude/skills/'inference-serving-vllm' (other agents: ~/.cursor/skills, ~/.copilot/skills, ~/.gemini/skills, ~/.gemini/config/skills, ~/.config/opencode/skills). Or install through the finder skill, which does both.
Badge for your README
The badge for davila7/claude-code-templates shows the score of its lowest-scoring skill.

[](https://zalavijaysinh.github.io/skilltrust/repos/davila7/claude-code-templates/)
Score history
| Date | Version | Score |
|---|
| 2026-11-10 | a6e6f93 | 70 |
| 2026-11-08 | b086f26 | 70 |
| 2026-11-07 | 395f30e | 70 |