
Linux System Troubleshooting, Performance Tuning & Automation
Diagnose Linux issues, tune performance, and automate system remediation with Claude
What You Can Do
This skill guides you through systematic Linux troubleshooting using Claude to analyze system logs, performance metrics, and configuration files. You'll identify root causes of system issues, optimize resource usage, and automate remediation tasks. Claude helps you parse complex diagnostic output and recommend fixes you can apply immediately.
Features
Feed Claude system logs (syslog, journalctl, application logs) to identify errors, warnings, and patterns in seconds
Analyze `top`, `iostat`, `vmstat`, and `netstat` output to pinpoint CPU, memory, I/O, and network bottlenecks
Review system configs (nginx, PostgreSQL, systemd, SSH) for misconfigurations and security issues
Generate shell scripts for automatic fixes you can review and deploy
Understand what high load averages, context switches, and memory pressure actually mean for your system
Analyze historical metrics to forecast resource needs and prevent outages
Compare configurations or metrics across multiple servers to detect drift and inconsistencies
Review system settings and suggest Linux hardening practices
Example Output
Input: CPU-heavy process consuming resources
User provides top output showing PostgreSQL at 85% CPU with high context switches.
Claude Output:
- Issue: PostgreSQL queries missing indexes or inefficient execution plans
- Evidence: Context switch rate (242/sec) indicates process scheduler contention
- Actions:
- Run EXPLAIN ANALYZE on slow queries
- Create missing indexes for top queries
- Adjust PostgreSQL work_mem parameter from 4MB to 16MB
- Script: Auto-generated psql commands to identify missing indexes
Input: Disk usage alert on /var partition
User provides du -sh /var/* output showing /var/log consuming 45GB.
Claude Output:
- Root cause: Verbose logging without rotation + 60-day retention
- Fix:
- Implement log rotation via
/etc/logrotate.d/app - Reduce log verbosity from DEBUG to WARN
- Compress archived logs with gzip
- Implement log rotation via
- Immediate recovery: Script to safely archive/compress existing logs
What's Included
- `SKILL.md`: Structured troubleshooting workflows, decision trees for common issues, log parsing templates
- `Linux Troubleshooting Decision Tree`: Interactive flowchart for diagnosing CPU, memory, disk, and network issues
- `Common Log Patterns Reference`: Regex patterns and interpretations for syslog, nginx, PostgreSQL, systemd-journal
- `Performance Metrics Cheat Sheet`: Quick reference for load average, context switches, iowait, and memory pressure
- `Remediation Script Generator Template`: Template prompts for Claude to generate safe, reviewable bash scripts
- `System Audit Checklist`: Security and configuration review checklist for hardening Linux systems
Who It's For
- Linux systems administrators managing production servers or clusters
- DevOps engineers diagnosing performance issues and automating remediation
- Database administrators analyzing query performance and resource contention
- Site reliability engineers (SREs) troubleshooting outages and optimizing infrastructure
- Full-stack developers deploying applications and investigating production issues
Best For
- Diagnosing unexplained spikes in CPU, memory, disk I/O, or network usage
- Analyzing application logs to identify errors, crashes, and performance degradation
- Optimizing database queries and system parameters for throughput and latency
- Automating routine system monitoring and remediation tasks with shell scripts
- Comparing configurations across multiple servers to detect drift and standardize settings





