Intel GPU AI Skills is a collection of agent skills for setting up, running, benchmarking, and profiling Hugging Face models on Intel GPUs. It supports workflows involving PyTorch, vLLM-XPU, SGLang-XPU, llama.cpp-SYCL, and migration from CUDA to XPU. The catalogue contains the project's skills, instructions, agent, and plugin.
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.
npx skills add intel/gpu-ai-skills --skill vllm-xpu-rungit clone --depth 1 https://github.com/intel/gpu-ai-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/skills/intel/gpu-ai-skills/vllm-xpu-run)<a href="https://agentmods.dev/skills/intel/gpu-ai-skills/vllm-xpu-run"><img src="https://agentmods.dev/badge/skills/intel/gpu-ai-skills/vllm-xpu-run/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/skills/intel/gpu-ai-skills/vllm-xpu-run"><img src="https://agentmods.dev/badge/skills/intel/gpu-ai-skills/vllm-xpu-run.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 6 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Privilege Escalation · line 52 Potential security issue detected. Manual review is recommended.Fix: Review the flagged content for security risks. Ensure no credentials, secrets, or sensitive data are exposed.
- high Privilege Escalation · line 75 Potential security issue detected. Manual review is recommended.Fix: Review the flagged content for security risks. Ensure no credentials, secrets, or sensitive data are exposed.
- high Privilege Escalation · line 184 Potential security issue detected. Manual review is recommended.Fix: Review the flagged content for security risks. Ensure no credentials, secrets, or sensitive data are exposed.
- high Privilege Escalation · line 72 Potential security issue detected. Manual review is recommended.Fix: Review the flagged content for security risks. Ensure no credentials, secrets, or sensitive data are exposed.
- medium MCP Rug Pull · line 71 Docker image references without a specific tag (:latest is implicit) or digest (@sha256:...) can be silently replaced by a malicious image.Fix: Pin the image: image:tag or image@sha256:abc123
- medium Data Exfiltration · line 95 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00129 | $0.03058 |
| Opus 5 | $0.00064 | $0.01529 |
| Sonnet 5 | $0.00026 | $0.00612 |
| Haiku 4.5 | $0.00013 | $0.00306 |
Grade A, and why
vllm-xpu-run scanned grade A with 1 finding 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 5d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -s http://localhost:8000/v1/chat/completions \ How it starts
The opening of the file, as written. The whole thing — 231 lines — stays where its author put it; the contents beside it link to each section on GitHub.
vllm-xpu-run
Use the official upstream vllm/vllm-openai-xpu:latest image.
Upstream vLLM has a first-class XPU backend. The CLI is identical
to the CUDA build (vllm serve <model>); device is detected from
torch.xpu.is_available(). There is no --device xpu flag.
The official image already sets ENTRYPOINT ["vllm", "serve"].
Pass the model id and serve flags directly after the image name. Do
not append another vllm serve: that produces
vllm serve vllm serve <model> and the container exits with code 2.
Current supported models and architectures
When the user asks which models vLLM supports on Intel XPU, fetch the current upstream page at request time:
https://docs.vllm.ai/en/stable/models/hardware_supported_models/xpu/
Do not answer from memory and do not copy a static model list into this skill. Report both the explicitly listed Model rows and the Architecture column, because the recommended model table is not an exhaustive checkpoint allowlist.
For a specific unlisted Hugging Face model, read its current
config.json and compare every value in architectures with the live
page's Architecture column. Report the evidence precisely:
- Exact model row → explicitly documented on the fetched page.
- Architecture match only → the architecture is documented on XPU, but this exact checkpoint is not explicitly validated by the page; perform a generation smoke test before claiming full support.
- Neither matches → not documented by the current XPU page; this is not proof of impossibility.
Include the source URL and retrieval date in the answer. If the page cannot be fetched, report that failure and offer to retry rather than substituting a remembered list. Do not infer XPU support merely from general vLLM, CUDA, or Transformers support.
CUDA → XPU cheat sheet
| CUDA convention | XPU convention |
|---|---|
vllm/vllm-openai:latest |
vllm/vllm-openai-xpu:latest |
--gpus all |
--device /dev/dri + -v /dev/dri/by-path:/dev/dri/by-path:ro + --ipc=host |
--dtype auto |
--dtype bfloat16 (explicit) |
| CUDA graphs default | --enforce-eager |
--tensor-parallel-size N |
same; pin N XPUs in ZE_AFFINITY_MASK |
What ships with it
4 files beside SKILL.md 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.
- 5d ago Changed 6924e9305c89
- 10d ago First seen · 231 lines · 129 tokens per session scan A 7310adc54c8a
vllm-xpu-run is a skill published in the GitHub repository intel/gpu-ai-skills (21 stars, last pushed 5d ago), licensed Apache-2.0. It adds 129 tokens to every session and 3,058 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
gemini-api-agent-platform
Guides the usage of the Gemini API on Agent Platform with the Google Gen AI SDK for enterprise AI applications. Covers SDK usage (Python, JS/TS, Go, Java, C#), capabilities like Live API, tools, multimedia generation, caching, and batch prediction.
open-source
Documentation reference for writing Python code using the browser-use open-source library. Use this skill whenever the user needs help with Agent, Browser, or Tools configuration, is writing code that imports from browseruse, asks about @sandbox deployment, supported LLM models, Actor API, custom tools, lifecycle…
deepstream-sop
Use this skill when building, deploying, evaluating, debugging, or measuring latency for the DeepStream SOP Inference Microservice — a GPU-accelerated FastAPI service that detects whether operators perform assembly-line steps in order via event boundary detection (GEBD) plus VLM classification. Trigger even if the…
gemini-api-dev
Use this skill when writing code that calls the Gemini API for text generation, multi-turn chat, multimodal understanding, image generation, video generation, streaming responses, background research tasks, function calling, structured output, or migrating from the old generateContent API. Covers SDK usage and best…
azure-search-documents-dotnet
Azure AI Search SDK for .NET (Azure.Search.Documents). Use for building search applications with full-text, vector, semantic, and hybrid search. Covers SearchClient (queries, document CRUD), SearchIndexClient (index management), and SearchIndexerClient (indexers, skillsets). Triggers: "Azure Search .NET"…
azure-search-documents-ts
Build search applications using Azure AI Search SDK for JavaScript (@azure/search-documents). Use when creating/managing indexes, implementing vector/hybrid search, semantic ranking, or building agentic retrieval with knowledge bases.