qzcli

Name: qzcli
Availability: InStock
Author: dogekiki

Manage GPU compute jobs on the Qizhi platform via command-line interface.

1|Updated Jul 21, 2026

Search Tags:#kubernetes #gpu #distributed-training #job-scheduling #qzcli #qizhi

Ready to run

npx skills add https://github.com/dogekiki/SP-test --skill qzcli-dogekiki

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: qzcli
Source: https://github.com/dogekiki/SP-test/tree/main/.trae/skills/qzcli
Command: npx skills add https://github.com/dogekiki/SP-test --skill qzcli-dogekiki

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires rich, requests, prompt_toolkit, mcp.

What problem does it solve?

This skill simplifies the complex process of managing GPU compute jobs on the Qizhi platform by providing a unified, kubectl-style interface that replaces manual web-portal interactions.

Core Features & Use Cases

Unified Job Management: Submit, monitor, and terminate distributed training jobs directly from your terminal.
Resource Discovery: Automatically cache and query available workspaces, compute groups, and hardware specifications.
Batch Processing: Execute large-scale training experiments using template-based batch submission configurations.

Quick Start

Use the qzcli skill to list all currently running jobs in your active workspace.

Frequently Asked Questions about qzcli

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I manage distributed training jobs on Kubernetes without using the web portal?▼

You can manage distributed training jobs on Kubernetes by using a unified command-line interface to submit, monitor, and terminate tasks directly from your terminal, bypassing manual web portal interactions entirely.

How do I submit GPU compute jobs for distributed training from the terminal?▼

Submit GPU compute jobs from the terminal by executing template-based batch submission configurations through a kubectl-style interface, facilitating large-scale training experiments and workspace discovery.

Do I need local configuration and authentication credentials to monitor GPU jobs?▼

Yes, monitoring GPU jobs requires local configuration of the command-line toolchain and authentication via credentials to successfully interact with the platform API and view real-time job statuses.

Can I execute batch processing for large-scale training experiments using command-line job scheduling?▼

Yes, you can execute large-scale training experiments through command-line job scheduling by utilizing template-based batch submission configurations to manage distributed training tasks efficiently.

What is the best way to discover available compute groups and hardware specifications for GPU jobs?▼

The best way to discover compute groups and hardware specifications is using command-line resource discovery, which automatically caches and queries available workspaces and hardware data for your distributed training jobs.