aiops-plugin

---

What is RHDP RCA Plugin?

RHDP RCA Plugin is a Claude Code marketplace containing specialized skills designed for Red Hat Demo Platform (RHDP) root cause analysis. This plugin suite enables AI-powered investigation of infrastructure failures, log analysis, and root cause diagnosis. These skills provide Claude with the tools to:

• Fetch and analyze logs from remote servers
• Correlate multiple data sources (Ansible, Splunk, GitHub)
• Perform automated root cause analysis
• Capture and organize user feedback

Quick Start

Prerequisites

• Claude Code installed
• SSH access to remote servers (for optional log auto-fetch)
• Splunk credentials (for log correlation)

1) Install the plugin

• Open Claude Code and run /plugin
• Add marketplace: redhat-et/rhdp-rca-plugin
• Install the plugin and restart Claude Code

2) Run root cause analysis

Start with a normal RCA request:

/aiops-plugin:root-cause-analysis job 123456

The root-cause-analysis workflow runs preflight checks and setup guidance first, then runs steps 1-4 automatically (log parse, Splunk correlation, GitHub context fetch), followed by Step 5 analysis.

3) Manual fallback (only if needed)

If preflight setup does not complete in your environment:

1. Copy and update .claude/settings.example.json.
2. Apply it to your local .claude/settings.local.json (project-level), including env vars and hooks.

Note: These hooks are required for MLflow tracing. The Stop hook flushes traces and the SessionStart hook captures the session ID.

For tracing:

• MLflow tracing is optional.
• If MLflow is still not configured after step 2, use MLflow Tracing Setup (Manual Fallback).

---

Available Skills

Skill	Description	Key Features
template-skill	Template for creating new skills	Starter template, best practices
logs-fetcher	Fetch Ansible/AAP logs via SSH	Time-based filtering, job number lookup
root-cause-analysis	Automated RCA for failed jobs	Log correlation, Splunk + GitHub integration
context-fetcher	Fetch job configs and docs	GitHub and Confluence integration
feedback-capture	Capture user feedback	Structured storage, categorization

---

Skill Details

🔍 logs-fetcher

Fetch Ansible/AAP logs from remote servers with flexible filtering

# Fetch logs from a specific time range
python -m scripts.fetch_logs_ssh \
  --start-time "2025-12-09 08:00:00" \
  --end-time "2025-12-10 17:00:00" \
  --mode processed

# Fetch logs by job number
python -m scripts.fetch_logs_by_job 1234567 1234568 1234569

Use cases:

• Fetch logs from specific time windows (minute/second precision)
• Retrieve logs for specific job numbers
• Download recent processed or ignored job logs
• Investigate incidents within a known timeframe

View detailed documentation →

---

🔎 root-cause-analysis

Investigate failed jobs by correlating Ansible/AAP logs with Splunk OCP pod logs and GitHub configuration

Step 1   [Python]  Parse local job log (extract GUID, namespace, failed tasks)
Step 2   [Python]  Query Splunk for correlated pod logs
Step 3   [Python]  Build correlation timeline
Step 4   [Python]  Fetch GitHub configs (AgnosticD/AgnosticV)
Step 5   [Claude]  Analyze and summarize root cause

Command Usage:

# By job ID (auto-fetches log from remote if not found locally)
.venv/bin/python scripts/cli.py analyze --job-id <JOB_ID> --fetch

# By explicit path (when you already have the log file)
.venv/bin/python scripts/cli.py analyze --job-log <path-to-job-log>

Use cases:

• Investigate job failures
• Analyze logs for errors and patterns
• Find root causes of infrastructure issues
• Debug failed deployments
• Troubleshoot Kubernetes/OpenShift problems