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

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

80%

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

Highly actionable and token-efficient CLI reference with strong copy-paste examples and a useful troubleshooting table. Its main weaknesses are the lack of an explicit validation checkpoint in the batch/destructive submission workflow and dangling references to bundle files that are not present.

Suggestions

Add an explicit post-submit verification step to the submission/batch workflow (e.g., run qzcli ls -r or qzcli status JOB_ID after submitting, and only report success once the job appears running).

Resolve or remove the references to ../shared-references/compute-env-contract.md and .aris/compute/qizhi.md — these files are not in the bundle, so the pointers are currently dangling.

Consider splitting the large per-subcommand reference into a separate REFERENCE.md and keeping SKILL.md as a concise overview pointing one level deep, which would improve progressive disclosure.

Dimension	Reasoning	Score
Conciseness	A dense reference-style manual with almost no prose padding and no explanation of concepts Claude already knows; every section is commands, flags, a parameter table, or a troubleshooting row that earns its place.	3 / 3
Actionability	Fully executable copy-paste bash commands throughout, a complete key-parameter table with defaults, a concrete batch_config.json example, and a troubleshooting table mapping problems to fixes.	3 / 3
Workflow Clarity	A numbered Quick Start sequence and dry-run previews exist, plus a troubleshooting feedback table, but batch/destructive submission lacks an explicit post-submit verification checkpoint (e.g., confirm via qzcli ls/status), capping this dimension at 2 per the rubric.	2 / 3
Progressive Disclosure	Sections are well organized, but the ~300-line body is monolithic with everything inline, and it references absent bundle paths (../shared-references/compute-env-contract.md, .aris/compute/qizhi.md) that do not exist in the bundle.	2 / 3
	Total	10 / 12 Passed

Description

100%

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 states the capability, names the tool and platform, and provides an explicit, well-chosen set of natural trigger phrases including bilingual terms. It clearly answers both what the skill does and when to use it.

Dimension	Reasoning	Score
Specificity	Names the concrete tool (qzcli) and platform (Qizhi/启智) and enumerates multiple specific concrete actions — "submit job", "stop job", "list jobs", "avail", "batch submit" — rather than vague language.	3 / 3
Completeness	Explicitly answers both what ("Manage GPU compute jobs on the Qizhi platform using qzcli — a kubectl-style CLI tool") and when ("Use when user says ... or needs to manage distributed training jobs on a Qizhi instance").	3 / 3
Trigger Term Quality	Strong natural-language coverage a user would actually say: "qzcli", "启智平台", "submit job", "stop job", "查计算组", "avail", "list jobs", "batch submit" — including bilingual platform terms.	3 / 3
Distinctiveness Conflict Risk	Tightly scoped to a single named platform and tool with bilingual distinctive triggers; very unlikely to fire for an unrelated skill.	3 / 3
	Total	12 / 12 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
Commit: 82076e5

Reviewed: 5 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.