TROUBLESHOOTING_SKILL

A troubleshooting procedure for etcd on two-node OpenShift clusters that use Pacemaker fencing, a system for isolating failed machines.

In plain words
What is it for?
Use it to validate Ansible and OpenShift access, gather Pacemaker and etcd diagnostics, investigate failures, perform remediation, and confirm whether the issue is resolved.
Why use it?
It gives the agent a structured way to check access, collect evidence, find causes, apply approved fixes, and verify the result.

Command for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/openshift-eng/two-node-toolbox/troubleshooting_skill
Clone the repo
git clone --depth 1 https://github.com/openshift-eng/two-node-toolbox

Made for: Claude Code.

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 5,275 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.05275
Opus 5 $0.00000 $0.02638
Sonnet 5 $0.00000 $0.01055
Haiku 4.5 $0.00000 $0.00528

Measured 2d ago against content hash ce1f62f6c0ed, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

TROUBLESHOOTING_SKILL scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Asks for rootmediumPrivilege escalation

A mod that escalates privileges can change anything on the machine, not only the project.

sudo pcs status
.claude/commands/etcd/TROUBLESHOOTING_SKILL.md · 603 lines

How it starts

The opening of the file, as written. The whole thing — 603 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Etcd Troubleshooting Skill

This document defines the Claude Code skill for troubleshooting etcd issues on two-node OpenShift clusters with fencing topology. When activated, Claude becomes an expert etcd/Pacemaker troubleshooter capable of iterative diagnosis and remediation.

Skill Overview

This skill enables Claude to:

  • Validate and test access to cluster components via Ansible and OpenShift CLI
  • Iteratively collect diagnostic data from Pacemaker, etcd, and OpenShift
  • Analyze symptoms and identify root causes
  • Propose and execute remediation steps
  • Verify fixes and adjust approach based on results
  • Provide comprehensive troubleshooting throughout the diagnostic process

Step-by-Step Procedure

1. Validate Access

1.1 Ansible Inventory Validation:

  • Check if deploy/openshift-clusters/inventory.ini exists
  • Verify the inventory file has valid cluster node entries
  • Test SSH connectivity to cluster nodes using Ansible ping module

1.2 OpenShift Cluster Access Validation:

  • Test direct cluster access with oc version
  • If direct access fails, check for deploy/openshift-clusters/proxy.env
  • If proxy.env exists, source it before running oc commands
  • Verify cluster access with oc get nodes
  • Remember proxy requirement for all subsequent oc commands

IMPORTANT: No Cluster Access Scenario

If OpenShift cluster API access is unavailable (which is expected when etcd is down), all diagnostics and remediation must be performed via Ansible using direct VM access. The troubleshooting workflow remains fully functional using only:

  • Ansible ad-hoc commands to cluster VMs
  • Ansible playbooks for diagnostics collection
  • Direct SSH access to nodes via Ansible

When cluster access is unavailable:

  • ✓ You can still diagnose and fix etcd issues completely
  • ✓ All Pacemaker operations work via Ansible
  • ✓ All etcd container operations work via Ansible (podman commands)
  • ✓ All logs are accessible via Ansible (journalctl commands)
  • ✗ Cannot query OpenShift operators or cluster-level resources
  • ✗ Cannot use oc commands for verification (use Ansible equivalents instead)

Read the full file on GitHub · 603 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 603 lines · 0 tokens per session scan B ce1f62f6c0ed

Subscribe to this mod's changes

TROUBLESHOOTING_SKILL is a command published in the GitHub repository openshift-eng/two-node-toolbox (5 stars, last pushed 28d ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 5,275 tokens. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.