Real Skill packageSource verifiedClawHub registry

huawei-cloud-modelarts-vllm-ascend-model-deploy

指导在华为云 ModelArts 推理平台通过 vLLM-Ascend 推理框架部署开源大语言模型(GLM/Qwen/DeepSeek/Llama 等)为昇腾 NPU 在线推理服务的完整流程。覆盖:确认模型后按 vllm-ascend 社区文档与用户确认部署形式(单机部署/多机部署/PD分离部署 3 种形式,统一单部署单元承载),从 HuggingFace/ModelScope 下载模型权重并上传 OBS、从 Quay.io 拉取 vllm-ascend 镜像并推送 SWR、按确认的部署形式选用专用脚本模板(单机/多机/PD 三套,脚本内不做部署形式运行时分支)生成启动/探针/角色识别脚本并上传 OBS、按 ModelArts 平台适配要点(端口 8080、OBS 挂载、bash 数组、探针命令检查、超节点亲和组)通过新版在线服务部署(二选一: 方式A 调用 CreateInferService API(POST /v2/{project_id}/services)自动部署 / 方式B 在 ModelArts 控制台按一步步点击指导手动部署;不再创建 AI 应用——该模块已废弃)、通过 OpenAI 兼容 API 调用验证…

Identity and source

Publisher attributionhuaweicloudskillregistry owner unverified by skillvetai
Functional categoryDevOps, Cloud & Observabilityautomatically inferred · 61% rule confidence
Package forminstruction with code32 recorded files
Canonical sourceClawHub registryclawhub:huaweicloudskill:huawei-cloud-modelarts-vllm-ascend-model-deploy
Open canonical source ↗

Platform declarations

These states come from the source or distribution context. None of the entries below are SkillVetAI compatibility test results.

OpenClawnative officialProvenance: registry distribution

Independent structural checks

These checks parse the fixed package against dated platform rules. They do not execute the Skill or verify task behavior.

Claude Codepasses structure
Checker 0.1.0 · agent-skills-2026-08-13+claude-code-docs-2026-08-13 · 11/17/2026.claude/skills/huawei-cloud-modelarts-vllm-ascend-model-deploy

Runtime, accounts, dependencies, permissions, network behavior and task quality remain untested.

OpenAI Codexpasses structure
Checker 0.1.0 · agent-skills-2026-08-13+codex-docs-2026-08-13 · 11/17/2026.agents/skills/huawei-cloud-modelarts-vllm-ascend-model-deploy

Runtime, accounts, dependencies, permissions, network behavior and task quality remain untested.

OpenClawpasses structure
Checker 0.1.0 · agent-skills-2026-08-13+openclaw-docs-2026-08-13 · 11/17/2026skills/huawei-cloud-modelarts-vllm-ascend-model-deploy

Runtime, accounts, dependencies, permissions, network behavior and task quality remain untested.

Installation and inspection

This command is recorded from the source ecosystem and resolves the registry's latest release. The fixed release shown on this page should be inspected before adoption.

clawhub install @huaweicloudskill/huawei-cloud-modelarts-vllm-ascend-model-deploy
clawhub inspect @huaweicloudskill/huawei-cloud-modelarts-vllm-ascend-model-deploy --version 1.0.6

Security evidence

SkillVetAI static result: medium signal

This automated, non-executing scan is bound to this release hash. It is not a safety certification and may contain false positives or false negatives.

Status
completed
Coverage
full text content
Files
31 / 32 inspected as text
Checked
11/17/2026, 7:54:23 AM
Scanner
0.1.3
Policy
1.0.3
1 automated finding
mediumRemote installer is downloaded without an integrity checkreferences/cli-installation-guide.md:28 · confidence 70%curl -sSL https://cn-north-4-hdn-koocli.obs.cn-north-4.myhuaweicloud.com/cli/latest/hcloud_install.sh
6 inferred permission indicators
  • shell execution — automatically inferred
  • network access — automatically inferred
  • filesystem read — automatically inferred
  • filesystem write — automatically inferred
  • credential access — automatically inferred
  • external write action — automatically inferred
11 dependency and API indicators
  • api: 0.0.0.0
  • api: clawhub.ai
  • api: cn-north-4-hdn-koocli.obs.cn-north-4.myhuaweicloud.com
  • api: docs.vllm.ai
  • api: github.com
  • api: iam.cn-southwest-301.beta.myhuaweicloud.com
  • api: quay.io
  • api: support.huaweicloud.com
  • api: www.apache.org
  • api: www.hiascend.com
  • api: www.modelscope.cn
External clawhub result: suspicious

This is registry-supplied evidence for the recorded release, not an independent SkillVetAI scan. Check the canonical source for the full report, scanner versions, scope, and current moderation state.

Evidence checked
11/17/2026, 2:26:30 AM
Release binding
Matches this record
  • skillspector: suspicious
  • llm: suspicious

Recorded files

The catalog stores hashes and an inventory summary for change detection. It does not republish the package contents.

Package content hashsha256:d03cfad12eed9d06eb460cf9a46a3a735f2d2a15920f39e96fd5a193d8448b36
Show up to 30 recorded paths
  • references/acceptance-criteria.md
  • references/api-deployment-guide.md
  • references/artifact-check.md
  • references/ascend-hardware-matrix.md
  • references/cli-installation-guide.md
  • references/deployment-form-decision.md
  • references/hixlep-guide.md
  • references/iam-policies.md
  • references/modelarts-console-guide.md
  • references/troubleshooting.md
  • references/verification-method.md
  • references/vllm-ascend-model-doc-guide.md
  • scripts/a5_hixlep/fix_ub_endpoint.py
  • scripts/a5_hixlep/gen_route.sh
  • scripts/a5_hixlep/generate_ep.py
  • scripts/a5_hixlep/get_route_eid.py
  • scripts/a5_hixlep/unofficial_ascend_tools-0.0.7rc4-py3-none-any.whl
  • scripts/check_artifacts.py
  • scripts/create_infer_service.py
  • scripts/download_model.py
  • scripts/get_master_ip.py
  • scripts/load_balance_proxy_server_example.py
  • scripts/obs_upload.sh
  • scripts/pull_push_image.sh
  • scripts/test_inference.py
  • scripts/upload_code_to_obs.py
  • skill-card.md
  • SKILL.md
  • templates/health_check.sh
  • templates/start_vllm.multi_node.sh

Source changelog

Version 1.0.6 expands deployment guidance and support for ModelArts resource pools. - Clarifies flavor selection for dedicated and public resource pools, including proxy sizing for PD deployments. - Updates deployment guides, scripts, and startup templates. - Adds the unofficial Ascend tools wheel. - Removes `skill-card.md`.

Release security diff

mediumCompared fixed releases 1.0.5 and 1.0.6; 0 finding and 0 permission indicators were added.

Both fixed releases were scanned under the current scanner and policy, so finding, permission and dependency changes are available.

Change reasons and limitations
  • file surface changed
  • content hash changed

Observed release history

These older immutable releases were observed by prior successful syncs. They remain recorded when a newer release becomes current.

1.0.511/11/2026sha256:a10f487400a94659ec28429be686c9d3d58cefe5c12da5e6166a7ca222c9fb2b
1.0.311/9/2026sha256:1b72c9a9fd2d3a94887c9dc4bf859b0840244d32b62fefe55f140da042ed6246
1.0.110/19/2026sha256:55771f52af333cb7825714f2499a62d588862915f566d66e49cb45d5b29b1129
1.0.010/14/2026sha256:0c2573cca0cc712532aab3e7c10e4285b1f4ad340ffca4b746febb11b33d43a0