Cross-platform, hardware-adaptive system optimizer. Auto-detects macOS / Linux / WSL / Windows (MINGW/Cygwin/MSYS2) and CPU/RAM/disk/GPU profile, then picks the right cleanup strategy. Scans reclaimable disk space, memory pressure, runaway processes, startup bloat, network issues. CleanMyMac built into Claude Code.
The English summary is being prepared. Showing the original description.
Review the instructions the agent will receive before you install.
Before scanning, load:
cat ${CLAUDE_PLUGIN_DATA_DIR:-$HOME/.claude/plugins/data/ops-ops-marketplace}/preferences.json — read timezone for timestampsThe bin/ops-speedup binary is the single source of truth for probes AND actions. This skill's job is to:
bin/ops-speedup| Command | Purpose | Side effects |
|---|---|---|
ops-speedup | Visual banner + hardware summary | None |
ops-speedup --json | Quick JSON diagnostics (disk/mem/net only) | None |
ops-speedup --scan | Full parallel probe: disk + mem + CPU hogs + power hogs + GPU/ANE + startup | None |
ops-speedup --clean | Safe cleanup: caches, tmp, logs, demote daemons, DNS flush, kernel tune | Non-destructive |
ops-speedup --deep | --clean + Trash, DerivedData, simulators, animation cuts, launch-agent kill | Removes files |
ops-speedup --aggressive | --deep + unload launch agents, docker --volumes, stale node_modules (>14d), TCP BBR | Potentially breaking — confirm first |
All modes:
~/.ops-speedup/history.jsonl| Capability | macOS | Linux | WSL | Windows |
|---|---|---|---|---|
| Disk reclaimable scan | ✓ | ✓ | ✓ | limited |
| Memory + swap | ✓ | ✓ | ✓ | limited |
| CPU hog kill | ✓ | ✓ | ✓ | — |
| Power/Energy Impact | ✓ (top -stats power) | ✓ (powertop) | — | — |
| GPU/Neural Engine | ✓ (powermetrics) | ✓ (nvidia-smi) | — | — |
| Launch agent offenders | ✓ | — | — | — |
| systemd unit masking | — | ✓ | ✓ | — |
| E-core demotion | ✓ (taskpolicy -b) | ✓ (renice+ionice) | ✓ | — |
| UI animation cuts | ✓ | — | — | — |
| Kernel tune (vnodes/somaxconn) | ✓ | ✓ | ✓ | — |
| TCP BBR | — | ✓ (aggressive) | ✓ (aggressive) | — |
| DNS flush | ✓ (dscacheutil) | ✓ (resolved) | ✓ (via Windows) | — |
| Memory purge | ✓ (purge) | ✓ (drop_caches) | ✓ | — |
| Stale build dir prune (>14d) | ✓ | ✓ | ✓ | — |
${CLAUDE_PLUGIN_ROOT}/bin/ops-speedup 2>/dev/null || echo "SCAN_FAILED"
${CLAUDE_PLUGIN_ROOT}/bin/ops-speedup --scan 2>/dev/null || echo '{}'
The binary already runs all probes in parallel. Do NOT add additional serial probe calls from this skill — they will duplicate work that's already in the JSON output.
Parse the JSON and render:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
OPS > SYSTEM SPEEDUP — [os] [os_version] [chip]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
HEALTH SCORE: [0-100] / 100 [████████░░ 80%]
DISK RECLAIMABLE
────────────────────────────────────────────────────
brew cache [N] MB ✓ safe
npm cache [N] MB ✓ safe
pnpm cache [N] MB ✓ safe
Xcode DerivedData [N] MB ✓ safe
Xcode DeviceSupport [N] MB ✓ safe
Docker reclaimable [N] MB ✓ safe
Metal shader cache [N] MB ✓ safe
Trash [N] MB ✓ safe
Logs [N] MB ✓ safe
Downloads [N] MB ⚠ review
Caches (general) [N] MB ⚠ review
/tmp [N] MB ✓ safe
apt/journal [N] MB ✓ safe (linux)
────────────────────────────────────────────────────
TOTAL RECLAIMABLE: [N] GB
MEMORY
────────────────────────────────────────────────────
Pressure: [N]% free Swap: [N] MB Free: [N] MB
NETWORK
────────────────────────────────────────────────────
Interface: [iface] DNS: [N]ms
STARTUP
────────────────────────────────────────────────────
Login items: [N] (macOS)
Launch agents: [N] (macOS)
Failed units: [N] (Linux)
Enabled units: [N] (Linux)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Health score calculation:
10 launch agents (macOS) or > 3 failed systemd units (Linux): -5
5GB reclaimable: -10
10GB reclaimable: -20
AskUserQuestion call 1 — Cleanup scope:
[Quick — caches, tmp, logs, DNS flush (~[N] GB)]
[Deep — + Trash, DerivedData, simulators, animation cuts (~[N] GB)]
[Aggressive — + launch-agent unload, stale node_modules, docker volumes (~[N] GB)]
[More options...]
AskUserQuestion call 2 (only if "More options..."):
[Custom — pick categories]
[Memory — kill top RAM hogs]
[Startup / Network / Skip...]
AskUserQuestion call 3 (only if "Startup / Network / Skip..."):
[Startup — review & disable launch agents / systemd units]
[Network — flush DNS, tune TCP, BBR (aggressive)]
[Skip — just show the report]
Per plugin Rule 5, destructive actions require explicit per-action confirmation. Before running --aggressive:
About to run AGGRESSIVE cleanup. Each item is destructive:
• Unload launch agents: [list]
• Docker volume prune (may delete unmounted volumes): [N] MB
• Stale node_modules (>14 days): [list of paths]
• TCP congestion control → BBR (Linux only)
[Proceed with all] [Pick categories] [Cancel]
If "Pick categories", batch per Rule 1 (max 4 options per AskUserQuestion).
Invoke the binary directly — it handles OS detection and dispatch:
# Quick clean
${CLAUDE_PLUGIN_ROOT}/bin/ops-speedup --clean
# Deep clean
${CLAUDE_PLUGIN_ROOT}/bin/ops-speedup --deep
# Aggressive (after confirmation)
${CLAUDE_PLUGIN_ROOT}/bin/ops-speedup --aggressive
Memory hog killing (option 5 from Phase 4):
Top processes are already in the scan JSON (cpu_hogs / power_hogs). Present them in paginated AskUserQuestion calls (max 3 processes + [More...] per page, final page has [Kill selected] + [Skip]).
NEVER kill: kernel_task, launchd, WindowServer, loginwindow, Finder, Dock, systemd, init, shells (bash/zsh/fish), tmux, IDE processes (Cursor/Comet/Code), Claude, node, python, Xcode. The binary's PROTECTED_RE regex blocks these automatically.
After cleanup, re-scan and diff:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
OPS > CLEANUP COMPLETE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Reclaimed: [N] GB
Disk free: [before] GB → [after] GB
RAM free: [before] MB → [after] MB
Swap: [before] MB → [after] MB
Health: [before]/100 → [after]/100
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
History: ~/.ops-speedup/history.jsonl
If user came from /ops:dash, offer b) Back to dashboard.
If $ARGUMENTS is:
scan or empty — Phase 1-3 only (report, no cleanup)clean — run ops-speedup --clean automatically (safe)deep — run ops-speedup --deep automatically (after 1 confirmation)auto — run ops-speedup --clean automatically, print resultsaggressive — run ops-speedup --aggressive after per-item confirmations~/.ops-speedup/history.jsonl is append-only. For trend questions ("is my disk filling up?", "is swap growing over time?"), read + graph:
tail -30 ~/.ops-speedup/history.jsonl | jq -r '[.ts, .ram_free_mb, .swap_mb, .disk_pct] | @tsv'
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Automate ActiveCampaign tasks via Rube MCP (Composio): manage contacts, tags, list subscriptions, automation enrollment, and tasks. Always search tools first for current schemas.
The English summary is being prepared. Showing the original description.
Pull the ads a competitor is running right now on Meta (Facebook/Instagram) and Google, then extract the hooks, offers, CTAs and landing pages worth modelling. Use when writing ad copy, planning creative tests, choosing a landing page angle, or when the user asks "what is X advertising", "what ads does X run", "find competitor ad hooks". Reads live Ad Library and Ads Transparency data through the Tracetify MCP; falls back to a manual Ad Library walkthrough when the connector is absent.
The English summary is being prepared. Showing the original description.
Analytics your AI agent can actually use. Track, analyze, run A/B experiments, and optimize across all your projects via CLI. Includes a growth playbook so your agent knows HOW to grow, not just what to track.
The English summary is being prepared. Showing the original description.
Teaches when to recall from long-term memory before acting and when to save durable decisions, corrections and failures afterwards. Use when a memory tool or MCP memory server is connected but the agent is not using it consistently, when the user complains that the assistant forgets preferences, conventions or past decisions between sessions, or when setting up persistent memory for a project. Works with any memory backend: a folder of Markdown files, a local MCP server, or a managed service.
The English summary is being prepared. Showing the original description.
Audit local AI coding-agent sessions with agenttrace for cost, tokens, tool failures, latency, anomalies, health, diffs, and CI gates.
The English summary is being prepared. Showing the original description.
Checks which private files (.env, keys, secrets) Claude Code and its subagents reached in this project, whether a value came into an agent's context, and which key to change, by running the open source agentwhy CLI over the session records Claude Code already keeps on disk. Use when the user asks whether the agent or a subagent read their .env or other secrets, which private files a session touched, whether a key leaked into the conversation, or which key to rotate. Runs locally and never prints a secret value.
The English summary is being prepared. Showing the original description.