Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/air-gapped/skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/plugins/air-gapped/skills/vllm)<a href="https://agentmods.dev/plugins/air-gapped/skills/vllm"><img src="https://agentmods.dev/badge/plugins/air-gapped/skills/vllm/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/plugins/air-gapped/skills/vllm"><img src="https://agentmods.dev/badge/plugins/air-gapped/skills/vllm.svg" alt="Reviewed on agentmods" width="80" height="20"></a>Grade A, and why
vllm scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
{
"name": "vllm",
"source": "./",
"description": "vLLM operator reference suite — deployment, configuration, quantization, caching, KV, tool parsers, reasoning parsers, chat templates, benchmarking, performance tuning, observability, omni, input modalities, speculative decoding, NVIDIA hardware, and a Gemma 4 31B operating-point serve recipe.",
"version": "0.20260820.85",
"author": {
"name": "Jörgen"
},
"license": "MIT",
"skills": [
"./.claude/skills/vllm-benchmarking",
"./.claude/skills/vllm-caching",
"./.claude/skills/vllm-chat-templates",
"./.claude/skills/vllm-configuration",
"./.claude/skills/vllm-deployment",
"./.claude/skills/vllm-gemma-4-31b",
"./.claude/skills/vllm-input-modalities",
"./.claude/skills/vllm-nvidia-hardware",
"./.claude/skills/vllm-observability",
"./.claude/skills/vllm-omni",
"./.claude/skills/vllm-performance-tuning",
"./.claude/skills/vllm-quantization",
"./.claude/skills/vllm-reasoning-parsers",
"./.claude/skills/vllm-speculative-decoding",
"./.claude/skills/vllm-tool-parsers"
],
"strict": false,
"category": "inference",
"tags": [
"vllm",
"llm",
"inference",
"kubernetes",
"nvidia",
"gpu",
"quantization",
"kv-cache"
]
}What it installs
The manifest is a name and a version. 15 skills travel with it, and installing the plugin installs all of them — 2,529 tokens a session between them. Each is measured on its own page, and each can be installed alone.
- Skill vllm-caching A 96 tokens
- Skill vllm-gemma-4-31b A 60 tokens
- Skill vllm-quantization A 226 tokens
- Skill vllm-speculative-decoding A 172 tokens
- Skill vllm-omni A 255 tokens
- Skill vllm-performance-tuning A 194 tokens
- Skill vllm-reasoning-parsers A 113 tokens
- Skill vllm-tool-parsers A 133 tokens
- Skill vllm-benchmarking A 126 tokens
- Skill vllm-configuration A 117 tokens
- Skill vllm-deployment A 272 tokens
- Skill vllm-observability A 174 tokens
- Skill vllm-chat-templates A 161 tokens
- Skill vllm-nvidia-hardware A 226 tokens
- Skill vllm-input-modalities B 204 tokens
What ships with it
1 file beside marketplace.json#vllm in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 39 lines scan A e98d6fec8ab7
vllm is a plugin published in the GitHub repository air-gapped/skills (5 stars, last pushed 9d ago), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other plugins, from other repositories
data-validation-suite
Schema validation, data quality monitoring, streaming validation pipelines, and input validation for backend APIs.
data
Data science and data engineering agents for visualization (D3, Highcharts), notebooks (Jupyter, Google Colab), R, and Python data engineering.
python-dev
Python development workflow with architecture design, coding guidelines, testing patterns, and specialized review agents.
branch-recap
Tells you what your branch actually changed — what was done, why, and what deserves a second look, instead of the diff — then files the fixes into the commits they belong to.
oatda
OATDA (One API To Direct All) — Unified access to 10+ LLM providers (OpenAI, Anthropic, Google, Deepseek, Mistral, xAI, Alibaba, MiniMax, ZAI, Moonshot, Bytedance) for text generation, vision analysis, image generation, video generation, audio generation/transcription/translation, model comparison, and balance…
data-engineering-skills
Data architecture, streaming pipelines, transformations, SQL optimization, and performance engineering.