debug-inference

Diagnose OpenShift AI InferenceService deployment issues via pod logs and GPU health.

48|31|Updated Feb 2, 2026

Search Tags:#troubleshooting #diagnosis #deployment issues #inference service #OpenShift AI

Ready to run

npx skills add https://github.com/RHEcosystemAppEng/agentic-plugins --skill debug-inference-rhecosystemappeng

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: debug-inference
Source: https://github.com/RHEcosystemAppEng/agentic-plugins/tree/main/rh-ai-engineer/skills/debug-inference
Command: npx skills add https://github.com/RHEcosystemAppEng/agentic-plugins --skill debug-inference-rhecosystemappeng

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires resources_get, resources_list, pods_list, pods_log, events_list, list_inference_services, get_inference_service, get_model_endpoint, get_deployment_info, analyze_vllm, chat_vllm, get_gpu_info, analyze_openshift, query_tempo_tool, get_trace_details_tool, execute_promql, korrel8r_get_correlated, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill is designed to help diagnose and resolve issues with InferenceService deployments on OpenShift AI.

Core Features & Use Cases

Progressive Diagnostics: Offers a step-by-step approach to identify and fix problems such as failed starts, slow latency, or errors in inference service deployments.
Model Deployment Issues: Assists with common problems including deployment delays, endpoint errors, slow performance, and GPU scheduling issues.
Observability and Analysis: Provides in-depth analysis of system status, events, pod logs, GPU health, and observability insights.
Human-in-the-Loop: Ensures a secure and responsible approach by requiring human confirmation for critical actions.

Quick Start

Run the /debug-inference skill to start diagnosing the issues with your InferenceService.

Frequently Asked Questions about debug-inference

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

FAQPage Schema

How do I troubleshoot inference service deployment issues on OpenShift AI?▼

Diagnosing slow latency in inference service deployments involves analyzing pod logs, GPU health, and observability insights using tools like query_tempo_tool and execute_promql to trace and resolve performance bottlenecks.

What is the best way to diagnose GPU scheduling issues for model deployments?▼

The best way to diagnose GPU scheduling issues for model deployments is to analyze GPU health and system events specifically within your OpenShift AI environment to identify resource constraints or scheduling failures.

Does this troubleshooting process require access to OpenShift AI cluster resources?▼

Yes, this troubleshooting process requires access to OpenShift AI cluster resources, utilizing multiple tools like pods_list, events_list, and get_gpu_info to perform in-depth diagnosis of your InferenceService deployments.

How do I analyze vLLM errors in my inference service?▼

To analyze vLLM errors in your inference service, use the analyze_vllm and chat_vllm tools to inspect model endpoints and deployment info, ensuring effective troubleshooting of runtime issues.

Why is my inference service failing to start on OpenShift AI?▼

Your inference service may be failing to start due to deployment delays, GPU scheduling issues, or configuration errors, which can be identified by analyzing system status, events, and pod logs.