qzcli

Manage GPU compute jobs on the Qizhi (启智) platform using qzcli — a kubectl-style CLI tool. Use when user says "qzcli", "启智平台", "submit job", "stop job", "查计算组", "avail", "list jobs", "batch submit", or needs to manage distributed training jobs on a Qizhi instance.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

High

Do not use without reviewing

Quality

Content

72%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.

The content is a strong, highly actionable CLI reference with executable examples throughout, but it is held back by a monolithic single-file structure, an opaque environment-contract preamble, and the absence of explicit validation feedback loops for batch and destructive operations.

Suggestions

Add an explicit validation checkpoint before destructive/batch operations (e.g. 'Run `qzcli create --dry-run` / `qzcli batch --dry-run`; only proceed if the preview is correct') and a confirm-before-stop note to lift workflow clarity above the cap.

Rewrite or trim the 'Environment contract' paragraph to remove unexplained jargon ("kernel witness", ".aris/compute/qizhi.md ledger") so it earns its tokens for a reader unfamiliar with the shared-references convention.

Split the long reference into one-level-deep bundle files (e.g. references/parameters.md, references/batch.md, references/hpc.md) and link to them from a concise overview to improve progressive disclosure.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient command-plus-comment reference material with parameter tables and minimal conceptual padding, but the opaque "Environment contract" paragraph (e.g. "kernel witness", ".aris/compute/qizhi.md ledger") adds jargon that could be trimmed or clarified, fitting score 4 rather than 5.	4 / 5
Actionability	Copy-paste-ready bash commands cover login, resource discovery, availability checks, interactive/non-interactive job creation, HPC/Slurm, batch submission with a full JSON config example, and job management, matching the score-5 anchor for fully executable guidance across common cases.	5 / 5
Workflow Clarity	A numbered Quick Start sequence exists and dry-run previews are shown for create/batch, but there is no explicit validate-then-proceed feedback loop and the destructive `stop JOB_ID` has no confirmation checkpoint, so per the batch/destructive cap workflow clarity cannot exceed 3.	3 / 5
Progressive Disclosure	Section headers are well organized, but the file is a ~325-line monolithic CLI reference with all detail inlined and only a single buried external reference ("../shared-references/compute-env-contract.md"); content that could live in separate files (parameter tables, batch config, HPC) is inline, fitting score 3.	3 / 5
	Total	15 / 20 Passed

Description

95%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.

The description is excellent: third-person voice, concrete domain framing, explicit Use-when clause, and rich bilingual trigger terms. It cleanly answers both what the skill does and when to invoke it with minimal conflict risk.

Dimension	Reasoning	Score
Specificity	Names the domain ("Manage GPU compute jobs on the Qizhi platform using qzcli — a kubectl-style CLI tool") and lists several specific actions via triggers ("submit job", "stop job", "list jobs", "batch submit"), with only minor coverage gaps, fitting the score-4 anchor rather than 5 because the actions are conveyed through trigger phrases instead of a dedicated capability list.	4 / 5
Completeness	Explicitly answers both what ("Manage GPU compute jobs on the Qizhi platform using qzcli") and when ("Use when user says ... or needs to manage distributed training jobs on a Qizhi instance"), matching the score-5 anchor with concrete trigger phrases.	5 / 5
Trigger Term Quality	Comprehensive natural trigger coverage including synonyms and bilingual phrasing ("qzcli", "启智平台", "submit job", "stop job", "查计算组", "avail", "list jobs", "batch submit"), matching the score-5 anchor for natural terms users would actually say.	5 / 5
Distinctiveness Conflict Risk	The Qizhi/qzcli niche with platform-specific bilingual triggers ("启智平台", "查计算组") is highly distinct and unlikely to fire for unrelated skills, matching the score-5 anchor.	5 / 5
	Total	19 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Path: skills/qzcli/SKILL.md
Commit: f4f20f9

Reviewed: 1 day 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.