minicpm5-deploy-vllm-ascend

Deploy MiniCPM5-2B with vLLM on Huawei Ascend NPU using vLLM-Ascend. Use when the user mentions vLLM-Ascend, Ascend NPU, Huawei Ascend, CANN, torch_npu, davinci devices, or wants an OpenAI-compatible MiniCPM5 server on Ascend hardware.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

82%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, highly actionable deployment guide with copy-paste commands, a sensible install→validate sequence, and a troubleshooting section. The only meaningful gap is that validation is implicit rather than framed as explicit feedback loops, plus one slightly narrative aside that could be trimmed.

Suggestions

Make the validation checkpoints explicit feedback loops, e.g. after the npu-smi/import check add 'If this fails, fix CANN/TorchNPU versions before continuing', and after the curl check add 'If not HTTP 200, consult Common failures'.

Trim the cookbook-provenance aside in the Install section (the sentence about parser responses produced on older vLLM with a compatibility shim) to keep the body leaner.

Consider splitting the multi-version Tool calling variant into a short reference note with a one-level link, since it duplicates much of the launch command with different version pins.

Dimension	Reasoning	Score
Conciseness	The body is lean and command-driven with no over-explanation of concepts Claude already knows, but the narrative aside about cookbook parser provenance and the compatibility shim is a minor instance of over-explanation that could be trimmed, fitting 'efficient; minor instances of over-explanation' rather than the fully lean anchor 5.	4 / 5
Actionability	Provides fully executable, copy-paste-ready docker, pip, environment-source, npu-smi, vllm-serve, and curl commands covering both the chat-completion and tool-calling cases, matching the 'fully executable; copy-paste ready; specific examples cover the common cases' anchor.	5 / 5
Workflow Clarity	Sections are clearly sequenced (Install → Prepare → Launch → Validate → Common failures) with a pre-launch driver/package check and a post-launch HTTP-200 expectation, but checkpoints are not framed as explicit validate→fix→retry loops, fitting 'clear sequence with most checkpoints present; minor validation gaps' rather than anchor 5.	4 / 5
Progressive Disclosure	No bundle files exist, but the body is well-organized into clearly headed sections with a one-level-deep Reference pointer to the cookbook plus an external guide, fitting 'good structure; most content appropriately placed; references mostly clear; minor organization gaps'.	4 / 5
	Total	17 / 20 Passed

Description

82%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, third-person description that cleanly answers both 'what' and 'when' with concrete, distinctive triggers for a well-scoped deployment niche. The only gap is breadth of action coverage, since it describes essentially one deploy action rather than several.

Dimension	Reasoning	Score
Specificity	Names a single concrete action ('Deploy MiniCPM5-2B with vLLM on Huawei Ascend NPU using vLLM-Ascend') with precise model/framework/hardware, but lists only one action rather than several, matching the '1-2 concrete actions, not comprehensive' anchor rather than the multi-action anchor 4.	3 / 5
Completeness	Explicitly states what it does ('Deploy MiniCPM5-2B with vLLM on Huawei Ascend NPU using vLLM-Ascend') and when to use it ('Use when the user mentions … or wants an OpenAI-compatible MiniCPM5 server on Ascend hardware'), matching the anchor for clear what-and-when with concrete triggers.	5 / 5
Trigger Term Quality	Lists natural user-facing terms ('vLLM-Ascend, Ascend NPU, Huawei Ascend, CANN, torch_npu, davinci devices') with good hardware-synonym coverage, fitting 'good keyword coverage; a few natural terms missing' rather than the fully comprehensive anchor 5.	4 / 5
Distinctiveness Conflict Risk	Targets a narrow niche (MiniCPM5-2B on Ascend via vLLM-Ascend) with hardware/software-specific triggers, giving a clear niche with minimal conflict risk.	5 / 5
	Total	17 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 1 suspicious	Warning

	Total	15 / 16 Passed

Repository: OpenBMB/MiniCPM
Path: skills/minicpm5-deploy-vllm-ascend/SKILL.md
Commit: 316cfb1

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.