main2main-error-analysis

Name: main2main-error-analysis
Availability: InStock
Author: nv-action

Diagnose vLLM-Ascend CI failures from upstream vLLM main branch changes.

4|6|Updated Feb 28, 2025

Search Tags:#automation #github #software-engineering #ci #root-cause-analysis #pull-request #log-analysis

Ready to run

npx skills add https://github.com/nv-action/vllm-benchmarks --skill main2main-error-analysis

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: main2main-error-analysis
Source: https://github.com/nv-action/vllm-benchmarks/tree/main/.agents/skills/main2main-error-analysis
Command: npx skills add https://github.com/nv-action/vllm-benchmarks --skill main2main-error-analysis

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Diagnoses and fixes vLLM-Ascend CI failures caused by upstream vLLM main branch changes. This skill implements a 4-phase pipeline: context acquisition, change analysis, report generation, and automated fix with PR submission.

Core Features & Use Cases

Interactive mode: given a CI run URL, extracts errors, generates report, applies fixes, creates PR
CI pipeline: given pre-extracted failures JSON, applies targeted fixes to existing branch
Produces a structured diagnostic report including per-run failures, root-cause errors, and proposed patches

Quick Start

Provide a CI run URL to analyze interactively, or supply a pre-extracted failures JSON to run in CI-pipeline mode.

Frequently Asked Questions about main2main-error-analysis

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

FAQPage Schema

How do I automate root-cause analysis for vLLM CI failures caused by upstream changes?▼

Yes, this skill automatically applies proposed patches and creates pull requests after diagnosing vLLM CI failures. Following the change analysis and report generation phases, it submits automated fixes directly to your existing branch via the GitHub CLI.

Can I analyze pre-extracted CI failure logs in a pipeline instead of using an interactive URL?▼

The environment requires the gh CLI tool for fetching CI logs and analyzing commits. The pipeline enforces environment checks during its context acquisition phase to ensure the necessary GitHub CLI dependencies are present before diagnosing vLLM-Ascend failures.

What is the best way to diagnose and fix vLLM-Ascend CI pipeline failures?▼

The best way to diagnose vLLM-Ascend CI pipeline failures is using a four-phase process: context acquisition, change analysis, report generation, and automated patch submission. This structured approach handles both interactive CI runs and pre-extracted failure logs for remediation.

Why does my vLLM-Ascend CI build break after upstream vLLM main branch evolution?▼

When you cannot provide a CI run URL, you can supply a pre-extracted failures JSON file. This switches the tool to CI-pipeline mode, allowing it to apply targeted fixes to an existing branch without needing direct access to the interactive GitHub CI environment.