skills.rest
villadalmine
villadalmineCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

holmesgpt

Troubleshoot HolmesGPT deployments and Kubernetes cluster questions via in-cluster AI assistant workflows.

1|Updated Apr 3, 2026
Search Tags:#troubleshooting#kubernetes#observability#prometheus#holmesgpt#litellm
One-click install
npx skills add https://github.com/villadalmine/infra --skill holmesgpt
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: holmesgpt
Source: https://github.com/villadalmine/infra/tree/main/.agents/skills/holmesgpt
Command: npx skills add https://github.com/villadalmine/infra --skill holmesgpt

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you ask plain-English questions about a Kubernetes cluster and get actionable answers powered by HolmesGPT, so you can inspect health, logs, and metrics without manually stitching together multiple tools.

Core Features & Use Cases

  • In-cluster troubleshooting: Query pods, deployments, and services in the ai namespace through HolmesGPT.
  • Metrics-aware analysis: Use Prometheus-backed questions to investigate CPU, memory, latency, and error spikes.
  • Operational support: Validate the LiteLLM OpenAI-compatible backend, test tool calling, and diagnose Gateway API exposure.
  • Example use cases: Find unhealthy workloads, review recent failures, compare resource usage, or ask whether Grafana or other services are degraded.

Quick Start

Ask HolmesGPT a plain-English question about your cluster, such as which pods are unhealthy or which workloads are using the most CPU, and let it query the in-cluster backend and observability tools for you.

Frequently Asked Questions about holmesgpt

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

FAQPage Schema
How do I troubleshoot unhealthy Kubernetes pods using natural language?▼

Find unhealthy Kubernetes workloads by asking plain-English questions about your cluster. The Skill queries in-cluster tools and observability integrations to inspect pod health, review recent failures, and return actionable diagnostic answers.

Can I query Prometheus metrics for CPU and memory spikes through an AI assistant?▼

Yes, you can query Prometheus metrics for CPU and memory spikes through an in-cluster AI assistant. The Skill applies Prometheus-backed metric queries to investigate latency, error spikes, and resource usage across your workloads using plain-English prompts.

Do I need a LiteLLM backend to use HolmesGPT for cluster observability?▼

Yes, you need an OpenAI-compatible LiteLLM backend with tool calling support to use HolmesGPT for cluster observability. The Skill requires this backend to process plain-English questions and query in-cluster Kubernetes tools effectively.

What is the best way to diagnose Gateway API exposure issues in the ai namespace?▼

The best way to diagnose Gateway API exposure issues in the ai namespace is to ask HolmesGPT directly. The Skill tests Gateway API access and validates your LiteLLM backend configuration to troubleshoot operational support questions within your cluster.

How does HolmesGPT investigate Kubernetes cluster questions without manual tool stitching?▼

HolmesGPT investigates Kubernetes cluster questions by applying in-cluster AI assistant workflows to pod inspection, log review, and metric queries. It uses tool calling support to automatically query observability integrations and return actionable answers.

Related Skills

Explore More Software Engineering Skills→
QianWen-AIQianWen-AI
67

qianwen-text

Enable OpenAI-compatible chat, text generation, and tool calling for Qwen models.

Official
Intermediate
openeuler-mirroropeneuler-mirror
3

swap-thrashing-diagnosis

Diagnose Linux swap thrashing by correlating system metrics with kernel memory behavior.

Official
Advanced
david-driscolldavid-driscoll
1

prometheus

Query Prometheus metrics, alerts, and rules via kubectl exec or port-forward.

Community
Intermediate
doleval013doleval013

k8s-debug

Diagnose K3s cluster issues with kubectl commands and log inspection.

Community
Basic