From system-check
Quick health check of the local development environment — memory, load, disk, Claude processes — and auto-invoke cleanup-claude-procs if signs of leftover Claude sessions are detected.
How this skill is triggered — by the user, by Claude, or both
Slash command
/system-check:system-checkThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Give a fast, readable snapshot of the development environment's health and flag whether any remediation is needed. Combines standard Linux commands (`free`, `uptime`, `df`, `ps`) with a targeted look for leftover Claude-spawned processes, and decides on its own whether to escalate to `/cleanup-claude-procs`.
Give a fast, readable snapshot of the development environment's health and flag whether any remediation is needed. Combines standard Linux commands (free, uptime, df, ps) with a targeted look for leftover Claude-spawned processes, and decides on its own whether to escalate to /cleanup-claude-procs.
free -h
uptime
df -h / /tmp 2>/dev/null | grep -v tmpfs
# Top 5 memory consumers
ps -eo pid,user,%mem,%cpu,etimes,command --sort=-%mem | head -6
# Claude-family process count (excluding this session's tree)
pgrep -af 'claude-code|@anthropic-ai/claude-code|playwright-mcp|chromium.*--headless' | wc -l
# Load vs. core count
nproc
# Local IP address
hostname -I | awk '{print $1}'
Evaluate each signal and classify OK / WARN / ALERT:
| Signal | OK | WARN | ALERT |
|---|---|---|---|
| Available memory | > 25% | 10–25% | < 10% |
| Swap used | < 25% | 25–60% | > 60% with free RAM < 20% |
| Load avg (1-min) / cores | < 0.7 | 0.7–1.5 | > 1.5 |
| Disk / usage | < 80% | 80–90% | > 90% |
| Stray claude/playwright procs (ppid=1) | 0 | 1–2 | 3+ |
| Uptime | any | > 30 days | n/a |
/cleanup-claude-procs (dry-run first, then confirm with user before killing)./cleanup-claude-procs, don't auto-run.$HOME and /tmp with du -sh (top 5).Use a compact, scannable format:
SYSTEM CHECK — 2026-04-15 09:31
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
IP: 192.168.1.100
Memory: 3.0 GiB available / 4.0 GiB total [OK]
Swap: 640 MiB / 2.0 GiB used [OK]
Load: 0.23 0.10 0.03 (4 cores) [OK]
Disk /: 42% used [OK]
Uptime: 1 day, 11h
Stray claude procs: 0 [OK]
→ All green. Nothing to do.
For a WARN/ALERT, add a Recommendations section with the specific next action and offer to run it.
Depending on findings, suggest:
/cleanup-claude-procs — for stray processesdu -sh ~/* | sort -h — for disk pressuresudo reboot — if uptime > 30 days AND swap pressure high/oh-my-claudecode:system-check
/oh-my-claudecode:system-check --verbose
/cleanup-claude-procs still goes through that skill's own confirmation flow.pgrep shows Claude procs but they all belong to the current session (check PPID chain via pstree), don't count them as stray.--verbose) may add per-process detail and full df output.Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub artificiallysocial/marketplace --plugin system-check