Requirements
- Target platform
- OpenClaw
- Install method
- Manual import
- Extraction
- Extract archive
- Prerequisites
- OpenClaw
- Primary doc
- SKILL.md
Manage Linux servers with user administration, process control, storage, and system maintenance.
Manage Linux servers with user administration, process control, storage, and system maintenance.
Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.
I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete.
I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run.
Create service accounts with --system flag โ no home directory, no login shell sudo with specific commands, not blanket ALL โ principle of least privilege Lock accounts instead of deleting: usermod -L โ preserves audit trail and file ownership SSH keys in ~/.ssh/authorized_keys with restrictive permissions โ 600 for file, 700 for directory visudo to edit sudoers โ catches syntax errors before saving, prevents lockout
systemctl for services, not service โ systemd is standard on modern distros journalctl -u service -f for live logs โ more powerful than tail on log files nice and ionice for background tasks โ don't compete with production workloads Kill signals: SIGTERM (15) first, SIGKILL (9) last resort โ SIGKILL doesn't allow cleanup nohup or screen/tmux for long-running commands โ SSH disconnect kills regular processes
df -h for disk usage, du -sh * to find culprits โ check before disk fills completely lsof +D /path finds processes using a directory โ needed before unmounting ncdu for interactive disk usage โ faster than repeated du commands Mount options matter: noexec, nosuid for security on data partitions Resize filesystems with care: grow is safe, shrink risks data loss โ always backup first
logrotate prevents disk fill โ configure size limits and retention Centralize logs to external system โ local logs lost if server dies /var/log/auth.log or /var/log/secure for login attempts โ watch for brute force dmesg for kernel messages โ hardware errors, OOM kills appear here Monitor inode usage, not just disk space โ many small files exhaust inodes
chmod 600 for secrets, 640 for configs, 644 for public โ world-writable is almost never correct Sticky bit on shared directories (chmod +t) โ users can only delete their own files setfacl for complex permissions โ when traditional owner/group/other isn't enough chattr +i makes files immutable โ even root can't modify without removing flag SELinux/AppArmor in enforcing mode โ permissive logs but doesn't protect
apt update before apt upgrade โ upgrade without update uses stale package lists Unattended security updates: unattended-upgrades โ critical patches shouldn't wait Pin package versions in production โ unexpected upgrades cause unexpected outages Remove unused packages: apt autoremove โ reduces attack surface and disk usage Know your package manager: apt/yum/dnf/pacman โ commands differ, concepts similar
Test restores regularly โ backups that can't restore are worthless Include package lists and configs, not just data โ recreating environment is painful Offsite backups mandatory โ local backups don't survive disk failure or ransomware Backup before any risky change โ "I'll just quickly edit" famous last words Document restore procedure โ 3am disaster is wrong time to figure it out
top/htop for live view, vmstat for trends โ understand baseline before diagnosing iotop for disk I/O bottlenecks โ slow disk often blamed on CPU Load average: 1.0 per core is healthy โ consistently higher means queuing Swap usage isn't inherently bad โ but consistent swapping indicates memory shortage sar for historical data โ retroactively diagnose what happened during incident
ss -tulpn shows listening ports โ netstat is deprecated ip addr and ip route replace ifconfig and route โ learn the new tools Check both host firewall and cloud security groups โ traffic blocked at either level fails /etc/hosts for local overrides โ quick testing without DNS changes curl -v shows full connection details โ headers, timing, TLS handshake
Running services as root โ one exploit owns the system No monitoring until something breaks โ reactive is expensive Editing config without backup โ cp file file.bak takes two seconds Rebooting to "fix" issues โ masks the problem, it'll return Ignoring disk space warnings โ 100% full causes cascading failures Forgetting timezone configuration โ logs from different servers don't correlate
Code helpers, APIs, CLIs, browser automation, testing, and developer operations.
Largest current source with strong distribution and engagement signals.