
system-admin
by dsyorkd
SKILL.md
name: system-admin description: > Automate system administration tasks including managing services, monitoring system health, handling user accounts, configuring firewalls, managing backups, and diagnosing network issues. Use when managing servers, deploying services, troubleshooting infrastructure, performing security audits, or automating operations tasks.
System Administration Toolkit
This skill provides automated system administration capabilities through production-grade utility scripts. You can manage services, monitor resources, handle users, configure security, perform backups, and diagnose issues on Linux/Unix systems.
Core Capabilities
Service Management
Manage systemd/init services with scripts/manage_service.py:
# Check service status
python3 scripts/manage_service.py status nginx
# Restart service
python3 scripts/manage_service.py restart nginx
# Enable at boot
python3 scripts/manage_service.py enable nginx
Supported operations: start, stop, restart, status, enable, disable, is-active, is-enabled
System Monitoring
Monitor system resources with scripts/monitor_system.py:
# Check all resources
python3 scripts/monitor_system.py --all
# Memory usage only
python3 scripts/monitor_system.py --memory
# Set alert thresholds
python3 scripts/monitor_system.py --all --cpu-threshold 80 --memory-threshold 90
Monitors: CPU usage, memory, disk space, load average, network I/O
User Management
Manage user accounts with scripts/manage_users.py:
# Create user
python3 scripts/manage_users.py create alice --groups sudo,docker
# List all users
python3 scripts/manage_users.py list
# Delete user (keeps home directory)
python3 scripts/manage_users.py delete bob --keep-home
Operations: create, delete, modify, list users and groups
Network Diagnostics
Diagnose network issues with scripts/network_diagnostics.py:
# Check connectivity
python3 scripts/network_diagnostics.py ping 8.8.8.8
# Full network report
python3 scripts/network_diagnostics.py check-all
# Test specific port
python3 scripts/network_diagnostics.py port-check example.com 443
Features: ping, DNS resolution, port scanning, traceroute, interface info
Firewall Management
Configure firewall rules with scripts/configure_firewall.py:
# Allow port
python3 scripts/configure_firewall.py allow 8080 --protocol tcp
# Block IP
python3 scripts/configure_firewall.py block 192.168.1.100
# List rules
python3 scripts/configure_firewall.py list
Supports: ufw, firewalld, iptables
Backup Operations
Automate backups with scripts/backup_system.py:
# Backup directory
python3 scripts/backup_system.py create /var/www /backup/www
# Restore from backup
python3 scripts/backup_system.py restore /backup/www-2026-01-03.tar.gz /var/www
# List backups
python3 scripts/backup_system.py list /backup
Features: compression, verification, incremental backups, retention policies
Additional Tools
All scripts support --help for detailed usage:
- Update System:
scripts/update_system.py- Package updates and security patches - Disk Management:
scripts/disk_management.py- Partition info, usage reports, cleanup - Process Management:
scripts/process_management.py- Monitor and manage processes - Schedule Tasks:
scripts/schedule_tasks.py- Cron job management - Log Rotation:
scripts/log_rotation.py- Configure log rotation policies - Security Audit:
scripts/audit_compliance.py- Security scans and compliance checks - Incident Response:
scripts/incident_response.py- Automated incident handling - Remote Access:
scripts/remote_access.py- SSH and remote access configuration - Virtualization:
scripts/virtualization_management.py- VM and container management - Software Inventory:
scripts/software_inventory.py- Track installed software - Install Software:
scripts/install_software.py- Automated software installation
Usage Guidelines
Safety First
- Scripts require appropriate permissions (use sudo when needed)
- Destructive operations prompt for confirmation (bypass with
--yes) - Always backup critical data before major changes
- Test in non-production environments first
Common Patterns
Check before acting:
# Check service status before restarting
python3 scripts/manage_service.py status nginx
python3 scripts/manage_service.py restart nginx
Chain operations:
# Update system and restart services
python3 scripts/update_system.py --security-only
python3 scripts/manage_service.py restart nginx postgresql
Monitor after changes:
# Deploy and monitor
python3 scripts/manage_service.py restart app
python3 scripts/monitor_system.py --all
python3 scripts/network_diagnostics.py check-all
System Context Integration
This skill works with hooks that inject system information:
- Current OS and distribution
- Init system (systemd, launchctl, etc.)
- Running services
- Resource usage (CPU, memory, disk)
- Network configuration
- Accumulated knowledge from previous sessions
Use this context to provide accurate, system-specific commands without querying.
Output Formats
Most scripts support multiple output formats:
# Human-readable (default)
python3 scripts/monitor_system.py --all
# JSON for parsing
python3 scripts/monitor_system.py --all --output json
# CSV for reports
python3 scripts/monitor_system.py --all --output csv
Advanced Usage
For complete tool documentation and advanced scenarios:
- reference.md - Complete documentation for all scripts
- examples.md - Common scenarios and workflows
Script Execution Notes
All scripts:
- Use Python 3.6+ (check with
python3 --version) - Include comprehensive
--helpdocumentation - Support dry-run mode where applicable (
--dry-run) - Provide detailed logging (use
--verbosefor debug output) - Return appropriate exit codes (0 = success, non-zero = failure)
- Handle errors gracefully with informative messages
Quick Reference
| Task | Script | Example |
|---|---|---|
| Service status | manage_service.py | python3 scripts/manage_service.py status nginx |
| System health | monitor_system.py | python3 scripts/monitor_system.py --all |
| Create user | manage_users.py | python3 scripts/manage_users.py create alice |
| Network test | network_diagnostics.py | python3 scripts/network_diagnostics.py ping 8.8.8.8 |
| Open firewall port | configure_firewall.py | python3 scripts/configure_firewall.py allow 80 |
| Backup directory | backup_system.py | python3 scripts/backup_system.py create /var/www /backup |
| Update packages | update_system.py | python3 scripts/update_system.py --security-only |
| Check disk usage | disk_management.py | python3 scripts/disk_management.py usage |
| Manage processes | process_management.py | python3 scripts/process_management.py list |
Troubleshooting
Script not found:
# Ensure you're in the skill directory or use full path
cd /path/to/skills/system-admin
python3 scripts/manage_service.py --help
Permission denied:
# Use sudo for privileged operations
sudo python3 scripts/manage_service.py restart nginx
Command not available:
# Scripts check for required tools and report missing dependencies
# Install missing tools based on error messages
Architecture: Tools Own State
Critical Pattern: All scripts in this skill are the source of truth for system operations.
- Scripts EXECUTE operations and return structured output (JSON where possible)
- LLM ORCHESTRATES by calling scripts with appropriate parameters
- LLM READS output to understand results and report to user
- LLM NEVER runs system commands directly (e.g.,
systemctl,useradd,iptables)
This pattern ensures:
- Consistent error handling and logging
- Proper permission management
- Audit trails and safety checks
- Reliable, repeatable operations
Available Commands
All scripts have corresponding slash commands for user invocation:
| Command | Script | Purpose |
|---|---|---|
/manage-service | manage_service.py | Start, stop, restart services |
/monitor-system | monitor_system.py | Check CPU, memory, disk |
/manage-users | manage_users.py | Create, delete, modify users |
/network-diagnostics | network_diagnostics.py | Ping, DNS, port checks |
/update-system | update_system.py | Apply system updates |
/configure-firewall | configure_firewall.py | Manage firewall rules |
/audit-compliance | audit_compliance.py | CIS, PCI-DSS, HIPAA audits |
/incident-response | incident_response.py | Security incident handling |
/remote-access | remote_access.py | SSH configuration |
/backup-system | backup_system.py | Create and restore backups |
/disk-management | disk_management.py | Disk usage and cleanup |
/virtualization | virtualization_management.py | VM and container management |
/process-management | process_management.py | Process monitoring and control |
/schedule-tasks | schedule_tasks.py | Cron job management |
/log-rotation | log_rotation.py | Log rotation policies |
/software-inventory | software_inventory.py | Track installed packages |
/install-software | install_software.py | Install/remove packages |
Integration with Agent
When using the system-admin agent, it will:
- Parse your natural language request
- Identify the appropriate script(s) to use
- Execute the script (not direct system commands)
- Read the output to understand results
- Chain multiple operations when needed
- Provide rich output with diagnostics and recommendations
You don't need to remember script names - just describe what you want to accomplish.
See also:
System-Specific Context
This section contains knowledge learned from previous sessions. It helps the system-admin skill provide more context-aware assistance.
System Facts
The following system information has been learned during previous sessions:
Operating System:
{
"type": "darwin",
"version": "25.0.0",
"kernel": "Darwin",
"architecture": "arm64"
}
Services:
{
"test-service": {
"status": "running",
"description": "Example service for testing"
},
"nginx": {
"status": "active",
"port": 80
}
}
Network Configuration:
{
"hostname": "macbook-pro.local"
}
Installed Packages:
{
"jq": "1.6",
"bash": "5.0"
}
Recent Command Results:
[
{
"commands": [
"ls -la",
"uname -a"
],
"notable_outputs": [
"System is darwin/macOS"
],
"errors_resolved": [],
"timestamp": "2026-01-04T00:10:20Z"
}
]
User Preferences
The following preferences have been learned from previous interactions:
Preferred Tools:
[
"curl",
"git",
"htop",
"tmux",
"vim"
]
Workflow Preferences:
{
"test_before_deploy": true,
"use_strict_mode": true
}
Coding Style:
{
"indent": "2 spaces"
}
Domain Knowledge
The following patterns and practices have been learned:
Last updated: 2026-01-03 18:10:58
This section is automatically generated from context files in context/
Score
Total Score
Based on repository quality metrics
SKILL.mdファイルが含まれている
ライセンスが設定されている
100文字以上の説明がある
GitHub Stars 100以上
3ヶ月以内に更新がある
10回以上フォークされている
オープンIssueが50未満
プログラミング言語が設定されている
1つ以上のタグが設定されている
Reviews
Reviews coming soon