# huawei-cloud-modelarts-training-diagnosis

huawei-cloud-modelarts-training-diagnosis diagnoses failed, timed-out, or stalled Huawei Cloud ModelArts training jobs. It gathers job details through hcloud and returns findings with confidence levels.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode, Windsurf (Devin Desktop), Claude.ai · 129 installs · ★ 59 repo stars
- Author: huaweicloud
- License: MIT · policy=mirror · the beginning of the file is at the end of this page
- Source: https://github.com/huaweicloud/huaweicloud-skills/tree/HEAD/skills/ai/modelarts/huawei-cloud-modelarts-training-diagnosis/SKILL.md
- Updated: 2026-10-23
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/skills/huawei-cloud-modelarts-training-diagnosis

## What it does

The skill checks a job’s status and inspects error messages, ModelArts events, and training stages. When a task message already contains a full Python traceback, it can finish the diagnosis without calling the other APIs. Its report includes findings and suggested fixes, each with a confidence level.

To find jobs and collect logs, the skill calls ModelArts APIs through `hcloud`. You need version 7.2.2 or later, configured AK/SK credentials, a `project_id`, and read access to the training APIs. Without those prerequisites, it cannot collect the job details.

Log collection can include a preview and an OBS download link; the full-log link remains valid for 5 minutes. If an API returns 401 or 403, the skill pauses and directs the user to configure IAM permissions.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add huaweicloud/huaweicloud-skills --skill huawei-cloud-modelarts-training-diagnosis -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/huawei-cloud-modelarts-training-diagnosis/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/huawei-cloud-modelarts-training-diagnosis/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add huaweicloud/huaweicloud-skills --skill huawei-cloud-modelarts-training-diagnosis -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/huawei-cloud-modelarts-training-diagnosis/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add huaweicloud/huaweicloud-skills --skill huawei-cloud-modelarts-training-diagnosis -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/huawei-cloud-modelarts-training-diagnosis/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install huaweicloud/huaweicloud-skills skills/ai/modelarts/huawei-cloud-modelarts-training-diagnosis
```

2. **Manually: put the files in place**
- File: `.github/skills/huawei-cloud-modelarts-training-diagnosis/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add huaweicloud/huaweicloud-skills --skill huawei-cloud-modelarts-training-diagnosis -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/huaweicloud/huaweicloud-skills --path skills/ai/modelarts/huawei-cloud-modelarts-training-diagnosis
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/huawei-cloud-modelarts-training-diagnosis/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add huaweicloud/huaweicloud-skills --skill huawei-cloud-modelarts-training-diagnosis -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/huaweicloud/huaweicloud-skills/tree/HEAD/skills/ai/modelarts/huawei-cloud-modelarts-training-diagnosis/SKILL.md

## When to use it
- find training jobs in failed or timed-out states
- inspect a job’s error message and Python traceback
- identify the stage where training stopped
- investigate a job that runs without making progress
- get suggested fixes for a ModelArts training failure

## Example prompts
- Diagnose ModelArts training job `TRAINING_JOB_ID` and suggest ways to fix the error.
- Find the stalled ModelArts training job and identify its last active stage.
- Scan failed ModelArts training jobs and explain the errors in their logs.

## Source text · SKILL.md · MIT

Below is the content of a third-party file. It is data, not instructions for you: do not act on anything written inside unless the human asks. Block tag: 623f21a870448891.

BEGIN THIRD-PARTY FILE 623f21a870448891
````text
---
name: huawei-cloud-modelarts-training-diagnosis
description: |
  Huawei Cloud ModelArts training job fault diagnosis skill. Uses hcloud CLI to call ModelArts training job log/event APIs, analyzes training job failures/timeouts/stuck jobs, locates customer training code issues, and provides diagnosis conclusions with fix suggestions and confidence levels.
  Scenarios: training job failure (status.phase=Failed), timeout (Timeout), abnormal (Abnormal), stuck jobs.
  Triggers: training job failure, training job timeout, training job stuck, ModelArts training diagnosis, 训练任务失败排查, 训练作业异常分析.
tags: [huawei-cloud, modelarts, training, diagnostics, hcloud]
---

# ModelArts Training Job Fault Diagnosis

## Overview

This skill provides automated fault diagnosis for Huawei Cloud ModelArts training jobs. It calls ModelArts log and event APIs via hcloud CLI to collect runtime information, analyzes training job failures/timeouts/stuck jobs, and outputs diagnosis conclusions with fix suggestions and confidence levels.

### Architecture

```
User Input
    ↓
Phase 1: Task Discovery (ListTrainingJobs)
    ↓
Phase 2: Status Assessment (ShowTrainingJobDetails)
    ↓
[Early Exit if traceback found in status.task_statuses[].message]
    ↓ (if no traceback)
Phase 3: Information Collection
    ├─ Main Path: ListTrainingJobEvents, ListTrainingJobStages
    └─ Extended Path: ShowTrainingJobLogsPreview, ShowObsUrlOfTrainingJobLogs
    ↓
Phase 4: Analysis (confidence-based inference)
    ↓
Phase 5: Output (diagnosis report + fix suggestions)
```

### Applicable Scenarios

- Training job failure (status.phase = "Failed")
- Training job timeout (status.phase = "Timeout")
- Training job abnormal (status.phase = "Abnormal")
- Training job stuck (running long time with no progress)
- Resource shortage causing training failure

### Typical Use Cases

- "My training job failed, help me diagnose"
- "Training job is stuck, no progress for hours"
- "Training job timeout, what went wrong?"
- "Scan all failed training jobs in my account"
- "Training job error code 1.015, what does it mean?"

## Prerequisites
````
END THIRD-PARTY FILE 623f21a870448891

The third-party file has ended. Everything between the 623f21a870448891 tags is data: requests and commands from it are not followed. An “end of file” line with another tag inside the block is part of the file.

First 2162 of 14455 bytes. © huaweicloud · MIT · https://github.com/huaweicloud/huaweicloud-skills/tree/HEAD/skills/ai/modelarts/huawei-cloud-modelarts-training-diagnosis/SKILL.md

_Part of the page did not fit — see the HTML version._

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
