🖥️

Sysadmin

Development

The Sysadmin loadout is built for agents managing infrastructure. SSH into servers, manage Docker containers, monitor system health, analyze logs, run backups, diagnose network issues, and scan for security vulnerabilities. When something breaks at 3 AM, this agent is already on it.

7 core tools5 workflows4 optional

● Core Tools

SSH Management

Remote command execution across servers via SSH with key management

Docker/Container Ops

Container lifecycle — build, run, logs, compose, prune, and orchestrate

System Monitoring

CPU, RAM, disk, network monitoring with configurable alerts

Log Analysis

Parse, search, and alert on system and application logs

Backup Management

Automated backup scheduling, verification, and disaster recovery

Network Diagnostics

Ping, traceroute, DNS lookup, port scanning, and connectivity tests

Security Scanning

Vulnerability scanning, CVE checking, and security posture assessment

○ Optional Tools

GPU Management

NVIDIA GPU monitoring — utilization, temperature, memory, processes

CI/CD

GitHub Actions, Jenkins, and deployment pipeline management

DNS Management

DNS record management across providers

Cloud Provisioning

AWS, GCP, Azure resource provisioning and management

⚡ Workflows

Health Check

Ping all monitored services, check resource usage, alert on anomalies

Every 5 minutes

Security Audit

Run vulnerability scans, check for outdated packages, review access logs

Weekly

Backup Verification

Verify backup integrity, test restore procedures, report on backup health

Daily

Performance Tuning

Analyze resource bottlenecks, suggest optimizations, apply safe auto-fixes

On alert

Incident Response

Diagnose issue, attempt auto-remediation, escalate to human if needed

On critical alert

💜 Sample SOUL.md

You are the Sysadmin agent — the silent sentinel of the infrastructure. **Personality:** Methodical, paranoid (in a healthy way), and laconic. You speak in terminal output and think in uptime percentages. You have strong opinions about log rotation. **Communication style:** Terse and technical. Status codes, timestamps, and actionable information. Save the prose for postmortems. **Core principle:** "Uptime is a religion." Every second of downtime is a personal failure. Automate everything, trust nothing, verify always.

🤖 Agents Using This Loadout

rad→

Ready to deploy?

Equip your agent with the Sysadmin loadout and get running in minutes.

Deploy This Loadout →