Agent skill

Vllm

by Prism-Shadow in Prism-Shadow/penguin-harness

Deploy and serve LLMs with vLLM behind an OpenAI-compatible endpoint, with tool calling enabled for agent workloads.

Apache-2.0Auto-check passedAI & LLM Engineering

Install Vllm

skills CLI

$ npx skills add Prism-Shadow/penguin-harness --skill vllm -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI

$ gh skill install Prism-Shadow/penguin-harness vllm --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy

$ git clone --depth 1 https://github.com/Prism-Shadow/penguin-harness.git skills-src && mkdir -p .claude/skills && cp -r skills-src/plugins/model-development/skills/vllm .claude/skills/vllm && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Paste into Claude Code

Install the "vllm" agent skill from https://github.com/Prism-Shadow/penguin-harness/tree/main/plugins/model-development/skills/vllm into .claude/skills/vllm/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "vllm", then confirm the skill loads.

Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

$skill-installer

$skill-installer install https://github.com/Prism-Shadow/penguin-harness/tree/main/plugins/model-development/skills/vllm

Type this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.

skills CLI

$ npx skills add Prism-Shadow/penguin-harness --skill vllm -a codex

Project install goes to .agents/skills/; add -g for ~/.codex/skills/.

GitHub CLI

$ gh skill install Prism-Shadow/penguin-harness vllm --agent codex

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/Prism-Shadow/penguin-harness.git skills-src && mkdir -p .agents/skills && cp -r skills-src/plugins/model-development/skills/vllm .agents/skills/vllm && rm -rf skills-src

Use ~/.agents/skills/ instead of .agents/skills for a personal install.

Codex skills documentation · loads skills from .agents/skills/

Paste into Codex

Install the "vllm" agent skill from https://github.com/Prism-Shadow/penguin-harness/tree/main/plugins/model-development/skills/vllm into .agents/skills/vllm/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "vllm", then confirm the skill loads.

Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add Prism-Shadow/penguin-harness --skill vllm -a cursor

Project install goes to .agents/skills/; add -g for ~/.cursor/skills/.

GitHub CLI

$ gh skill install Prism-Shadow/penguin-harness vllm --agent cursor

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/Prism-Shadow/penguin-harness.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/plugins/model-development/skills/vllm .cursor/skills/vllm && rm -rf skills-src

Use ~/.cursor/skills/ instead of .cursor/skills for a personal install.

Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/

Paste into Cursor

Install the "vllm" agent skill from https://github.com/Prism-Shadow/penguin-harness/tree/main/plugins/model-development/skills/vllm into .cursor/skills/vllm/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "vllm", then confirm the skill loads.

Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

gemini skills install

$ gemini skills install https://github.com/Prism-Shadow/penguin-harness.git --path plugins/model-development/skills/vllm

--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.

skills CLI

$ npx skills add Prism-Shadow/penguin-harness --skill vllm -a gemini-cli

Project install goes to .agents/skills/; add -g for ~/.gemini/skills/.

GitHub CLI

$ gh skill install Prism-Shadow/penguin-harness vllm --agent gemini-cli

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/Prism-Shadow/penguin-harness.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/plugins/model-development/skills/vllm .gemini/skills/vllm && rm -rf skills-src

Use ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.

Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/

Paste into Gemini CLI

Install the "vllm" agent skill from https://github.com/Prism-Shadow/penguin-harness/tree/main/plugins/model-development/skills/vllm into .gemini/skills/vllm/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "vllm", then confirm the skill loads.

Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

GitHub CLI

$ gh skill install Prism-Shadow/penguin-harness vllm

Installs for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).

skills CLI

$ npx skills add Prism-Shadow/penguin-harness --skill vllm -a github-copilot

Project install goes to .agents/skills/; add -g for ~/.copilot/skills/.

Manual copy

$ git clone --depth 1 https://github.com/Prism-Shadow/penguin-harness.git skills-src && mkdir -p .github/skills && cp -r skills-src/plugins/model-development/skills/vllm .github/skills/vllm && rm -rf skills-src

Use ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.

GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/

Paste into GitHub Copilot

Install the "vllm" agent skill from https://github.com/Prism-Shadow/penguin-harness/tree/main/plugins/model-development/skills/vllm into .github/skills/vllm/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "vllm", then confirm the skill loads.

GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add Prism-Shadow/penguin-harness --skill vllm -a opencode

OpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.

GitHub CLI

$ gh skill install Prism-Shadow/penguin-harness vllm --agent opencode

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/Prism-Shadow/penguin-harness.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/plugins/model-development/skills/vllm .opencode/skills/vllm && rm -rf skills-src

Use ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.

OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/

Paste into OpenCode

Install the "vllm" agent skill from https://github.com/Prism-Shadow/penguin-harness/tree/main/plugins/model-development/skills/vllm into .opencode/skills/vllm/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "vllm", then confirm the skill loads.

OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

Facts

Skill name: vllm
GitHub stars: 2.5k
Token cost: ~1k tokens
SKILL.md length: 461 words
Files: 1
Skills in repo: 31
Repo updated: Oct 7, 2026
First seen: Oct 7, 2026
Licence: Apache-2.0

At a glance

Deploy and serve LLMs with vLLM behind an OpenAI-compatible endpoint, with tool calling enabled for agent workloads.

Works in 6 steps: Ask the user which model to serve; with… → Pick the engine the user prefers: vLLM… → Serve on a free port, with the… → …
Tasks that involve LLM inference and serving
SKILL.md covers Before you start, Suggested workflow, Install and Serve, plus 4 more sections
Calls python3, curl and pip

What it does

Vllm is an agent skill from Prism-Shadow/penguin-harness. Deploy and serve LLMs with vLLM behind an OpenAI-compatible endpoint, with tool calling enabled for agent workloads.

Its SKILL.md is about 1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in AI & LLM Engineering, covering LLM inference and serving and Structured output and tool calling. It works with vLLM, OpenAI, Qwen and Ollama. The repository describes itself as: 🐧 Unified and Stable RSI Platform. The licence is Apache-2.0.

When your agent uses it

Tasks that involve LLM inference and serving
Tasks that involve Structured output and tool calling

Example prompts

“/vllm”

Requirements

Python 3

Workflow steps

6 steps, taken from the first numbered list in SKILL.md.

Ask the user which model to serve; with no preference, recommend Qwen/Qwen3.5-0.8B.
Pick the engine the user prefers: vLLM for high-throughput GPU serving; Ollama is the simple default and the choice on macOS or CPU-only…
Serve on a free port, with the tool-calling flags whenever agents will call it (see below).
Verify with curl http://localhost:8000/v1/models.
Register the endpoint: penguin config model add ... --client-type openai-chat --base-url http://localhost:8000/v1 — a served model is not…
Confirm the new entry with penguin config model list.

What it can do on your machine

Read from SKILL.md and the folder at commit d56d9ce. It shows what the files ask for, not the result of running them.

Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Shell commands in SKILL.md call:
- python3
- curl
- pip
From the folder's file list and the shell code blocks in SKILL.md.
Network
Links to these hosts (documentation or services it may open):
- huggingface.co
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Vllm loads about 1k tokens when it runs. Until then it costs about 30 tokens; SKILL.md has 461 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it: ~30
When it runs · the whole SKILL.md, loaded when a task matches: ~1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from Prism-Shadow/penguin-harness at commit d56d9ce, republished under its Apache-2.0 licence (© Prism-Shadow). 461 words, ~1,026 tokens.

Download SKILL.mdSave it as .claude/skills/vllm/SKILL.md (or your agent's skills folder).

name: vllm
description: Deploy and serve LLMs with vLLM behind an OpenAI-compatible endpoint, with tool calling enabled for agent workloads.

vLLM Serving

vLLM serves open-weight LLMs on local GPUs with high-throughput inference behind an OpenAI-compatible API, ready for chat and agent workloads.

Before you start

If the user's message only invokes this skill (e.g. "use vllm skill") without a concrete request, ask the user what they want. Do not run any command until the goal is clear.

Ask the user which model to serve; if they have no preference, recommend the small default Qwen/Qwen3.5-0.8B. Also ask what context length the workload needs.

vLLM needs an NVIDIA or AMD GPU. Engine choice follows the user's preference: Ollama also runs on GPUs and is the simpler default — pick vLLM for high-throughput serving, and Ollama on macOS or CPU-only machines, which vLLM does not serve. Confirm the hardware first:

bash

nvidia-smi          # NVIDIA: GPU model and free VRAM (AMD ROCm: rocm-smi)
python3 --version   # a recent Python is required

The model must fit the available VRAM — model size and context length drive the serve flags below.

Suggested workflow

Ask the user which model to serve; with no preference, recommend Qwen/Qwen3.5-0.8B.
Pick the engine the user prefers: vLLM for high-throughput GPU serving; Ollama is the simple default and the choice on macOS or CPU-only machines.
Serve on a free port, with the tool-calling flags whenever agents will call it (see below).
Verify with curl http://localhost:8000/v1/models.
Register the endpoint: penguin config model add ... --client-type openai-chat --base-url http://localhost:8000/v1 — a served model is not visible to Penguin until added.
Confirm the new entry with penguin config model list.

Install

Use a fresh virtual environment (or uv):

bash

python3 -m venv .venv && source .venv/bin/activate
pip install vllm

Show full SKILL.md (216 more words)Show less

Serve

bash

vllm serve Qwen/Qwen3.5-0.8B --port 8000

This exposes an OpenAI-compatible API at http://localhost:8000/v1. Key flags:

--served-model-name <name> — the model id clients request (defaults to the model path).
--api-key <key> — require this bearer token on every request.
--max-model-len <n> — context window; agent sessions need a large one.
--gpu-memory-utilization <0..1> — fraction of VRAM to claim (default 0.9).
--tensor-parallel-size <n> — shard across n GPUs.
--dtype <auto|bfloat16|float16> and --quantization <awq|gptq|fp8> — precision and quantized weights.

If the port is taken, pick a free one — never kill a process already listening on it.

Tool calling — required for agents

Agent harnesses (PenguinHarness included) send tools with their requests. vLLM must opt in at startup:

bash

vllm serve Qwen/Qwen3.5-0.8B --enable-auto-tool-choice --tool-call-parser hermes

Choose the parser for the model family — e.g. hermes for Qwen models, llama3_json for Llama models. Without these flags, requests that set tool_choice fail with 400 "auto" tool choice requires --enable-auto-tool-choice and --tool-call-parser to be set.

Verify

bash

curl http://localhost:8000/v1/models

Register with PenguinHarness

Model configuration is the penguin CLI's job — penguin config model add registers an endpoint and penguin config model list shows what has been registered. A served model is not visible to Penguin until you add it:

bash

penguin config model add --provider custom --client-type openai-chat \
  --base-url http://localhost:8000/v1 --model-id <served-model-name> --api-key <key>
penguin config model list   # the new entry should now be listed

Troubleshooting

Out of memory at startup: lower --gpu-memory-utilization or --max-model-len, or serve a quantized model.
Long prompts truncated or context-length errors: raise --max-model-len (bounded by VRAM).
400 on tool calls: restart the server with the tool-calling flags above.

Files

Just SKILL.md in plugins/model-development/skills/vllm of Prism-Shadow/penguin-harness.

Open the folder on GitHubat commit d56d9ce

Compare with similar skills

Vllm next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Vllm compared with similar skills
Skill	Stars	Used in	Tokens	Auto-check	Licence	Repo updated
Vllm this skillPrism-Shadow/penguin-harness	2.5k	—	~1k	Automated safety check: Pass	Apache-2.0	today
Aider DelegateamElnagdy/delegate-skills	2.3k	3 repos	~3k	Automated safety check: Pass	MIT	today
Perfupraullenchai/Rapid-MLX	3.9k	—	~1.6k	Automated safety check: Notes	Custom licence	today
Resolvealexziskind1/model-shelf	130	—	~792	Automated safety check: Pass	MIT	1 mo ago
Tanstack AIsecondsky/claude-skills	227	1 repos	~3.6k	Automated safety check: Notes	MIT	9 days ago
Vllm Ascendascend-ai-coding/awesome-ascend-skills	174	—	~2.7k	Automated safety check: Pass	None	today

Similar skills

Aider Delegate
amElnagdy/delegate-skills
Delegate a coding task to Aider (aider) as a background implementer, then review its diff and land it yourself.
2.3k GitHub starsUsed in 3 repos~3k tokens
AI & LLM EngineeringAuto-check passed
Perfup
raullenchai/Rapid-MLX
Autonomous performance optimization: research, PoC, benchmark, implement, review, PR
3.9k GitHub stars~1.6k tokensUpdated today
AI & LLM EngineeringAuto-check: notes
Resolve
alexziskind1/model-shelf
Always resolve Hugging Face models via model-shelf before any download.
130 GitHub stars~792 tokensUpdated 1 mo ago
AI & LLM EngineeringAuto-check passed
Tanstack AI
secondsky/claude-skills
TanStack AI (alpha) provider-agnostic type-safe chat with streaming for OpenAI, Anthropic, Gemini, Ollama.
227 GitHub starsUsed in 1 repo~3.6k tokens
AI & LLM EngineeringAuto-check: notes
Vllm Ascend
ascend-ai-coding/awesome-ascend-skills
vLLM Ascend plugin for LLM inference serving on Huawei Ascend NPU.
174 GitHub stars~2.7k tokensUpdated today
AI & LLM EngineeringAuto-check passed
Vllm
majiayu000/claude-skill-registry
A skill your agent uses when self-hosting an open-weight LLM for high-throughput concurrent serving with vLLM — running an OpenAI-compatible endpoint, splitting a model across GPUs with tensor or…
666 GitHub starsUsed in 1 repo~3.6k tokens
AI & LLM EngineeringAuto-check passed

More from Prism-Shadow/penguin-harness

All 31 skills in this repo

A2ui
Prism-Shadow/penguin-harness
Make a reply easier to read and act on with rich blocks inside ordinary Markdown — a choice the user picks from, a form that collects several answers, a procedure as steps with warnings in place, a…
2.5k GitHub stars~3k tokensUpdated today
Auto-check passed
Penguin Harness Dev
Prism-Shadow/penguin-harness
A skill your agent uses when developing PenguinHarness itself — changing packages/{core,server,web,cli,desktop,landing,docs,skills}, the built-in model catalog, the installers or the release…
2.5k GitHub stars~3.4k tokensUpdated today
Auto-check passed
Bento Slides
Prism-Shadow/penguin-harness
Create and edit Bento presentations — self-contained .bento.html decks whose document is JSON.
2.5k GitHub stars~1.6k tokensUpdated today
Auto-check passed
Penguin Harness Manual Test
Prism-Shadow/penguin-harness
A skill your agent uses when standing PenguinHarness up to try a change by hand — launching the Web App, the desktop shell, the landing page, the docs site or the component gallery to click through…
2.5k GitHub stars~1.3k tokensUpdated today
Auto-check passed
Penguin Harness Frontend
Prism-Shadow/penguin-harness
A skill your agent uses when changing the PenguinHarness Web App (packages/web) or the shared UI package — adding or restyling any UI, picking a status colour, adding an icon, laying out a row or a…
2.5k GitHub stars~6.4k tokensUpdated today
Auto-check passed
Browser Automation
Prism-Shadow/penguin-harness
Drive the PenguinHarness agent browser — the desktop app's built-in browser or the user's own Chrome — from the shell with penguin browser: open pages, read them as simplified HTML or text, act with…
2.5k GitHub stars~2.9k tokensUpdated today
Auto-check: warnings

Works with

Questions about Vllm

What does Vllm do?

Deploy and serve LLMs with vLLM behind an OpenAI-compatible endpoint, with tool calling enabled for agent workloads. Vllm is an agent skill from Prism-Shadow/penguin-harness. Deploy and serve LLMs with vLLM behind an OpenAI-compatible endpoint, with tool calling enabled for agent workloads.

When should I use Vllm?

Vllm fits situations like: tasks that involve LLM inference and serving; tasks that involve Structured output and tool calling.

How do I install Vllm in Claude Code?

Run `npx skills add Prism-Shadow/penguin-harness --skill vllm -a claude-code`. Or copy the skill folder (plugins/model-development/skills/vllm in Prism-Shadow/penguin-harness) into .claude/skills/vllm in your project. Claude Code loads it when a task matches its description.

How do I install Vllm in Codex?

Run `npx skills add Prism-Shadow/penguin-harness --skill vllm -a codex`. Or copy the skill folder (plugins/model-development/skills/vllm in Prism-Shadow/penguin-harness) into .agents/skills/vllm in your project. Codex loads it when a task matches its description.

Can I use Vllm in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add Prism-Shadow/penguin-harness --skill vllm -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/vllm, .gemini/skills/vllm, .github/skills/vllm and .opencode/skills/vllm in your project.

What does Vllm need to run?

Going by SKILL.md and its folder, Vllm needs the command-line tools its instructions call (python3, curl and pip). Our summary lists: Python 3.

Does Vllm access the network?

SKILL.md names 1 domain. As links in the text: huggingface.co. This is read from the text; nothing was executed.

Is Vllm safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Vllm use?

Vllm is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Vllm use?

About 1k tokens (SKILL.md is roughly 4.1k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Vllm?

Skills that share tags, products or a category with Vllm: Aider Delegate (amElnagdy/delegate-skills, 2.3k stars), Perfup (raullenchai/Rapid-MLX, 3.9k stars), Resolve (alexziskind1/model-shelf, 130 stars) and Tanstack AI (secondsky/claude-skills, 227 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Vllm?

Prism-Shadow (a GitHub organization) maintains it in Prism-Shadow/penguin-harness, which has 2,455 GitHub stars. The repository holds 31 skills in this directory. The repository was last updated on October 7, 2026.

Source: Prism-Shadow/penguin-harness on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.

Install Vllm

Facts

At a glance

What it does

When your agent uses it

Example prompts

Requirements

Workflow steps

What it can do on your machine

Tool permissions

Runs code

Network

Credentials

Context cost

Safety

SKILL.md

vLLM Serving

Before you start

Suggested workflow

Install

Serve

Tool calling — required for agents

Verify

Register with PenguinHarness

Troubleshooting

Files

Compare with similar skills

Similar skills

More from Prism-Shadow/penguin-harness

Related

Questions about Vllm