本文へ移動
cccskills
無料GitHub で公開

sovereign-lockdown

Apply the full DSH sovereign lockdown: kill known phone-home daemons (Google, Zoom, CodeWhisperer), pin DNS to dnscrypt-proxy, enable firewall + stealth, install a pf anchor blocking telemetry endpoints, and permanently remove unauthorized LaunchAgents. Verifies with audit.sh + dome-check.sh and surfaces any remaining gaps.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md5.7 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

You are hardening the local machine into a sovereign state where nothing calls home, no unauthorized background daemon persists, and network traffic is policy-gated. This is invasive — every step needs the user's sudo once, and you MUST surface what each phase changes before running.

1. Confirm intent

Before touching the system, show the user what will happen:

Sovereign Lockdown will:
  • enable macOS firewall + stealth mode (blocks unsolicited inbound)
  • remove ~/Library/LaunchAgents entries not in the approved list
    (Google updater, Zoom daemon, CodeWhisperer, etc.)
  • pin DNS to 127.0.0.1 (dnscrypt-proxy must be running)
  • install a pf anchor blocking known telemetry endpoints
  • set pmset: no sudden-motion-sensor, no powernap, destroy FV key on standby
  • disable AirPlay Receiver (closes ports 5000/7000)

Require an explicit "yes" to proceed. If the user hesitates, offer to run only lock-down.sh (phase 1, least invasive) and skip phases 2–4.

2. Run phase 1 — kill daemons + pin DNS

bash "$DOME_ROOT/scripts/lock-down.sh"

Prompts for sudo once, keeps credential alive for the whole run. What it does:

  • launchctl unload -w + rm on unauthorized user LaunchAgents
  • launchctl bootout system on unauthorized system daemons
  • sudo networksetup -setdnsservers Wi-Fi 127.0.0.1
  • sudo brew services start dnscrypt-proxy
  • Kills any running processes of the daemons it just removed

Verify before proceeding:

pgrep -x dnscrypt-proxy >/dev/null && echo "✓ dnscrypt-proxy running" || echo "✗ MISSING"
scutil --dns | grep -m1 nameserver | awk '{print $3}'   # should be 127.0.0.1

3. Run phase 2 — firewall + stealth + hardware tuning

bash "$DOME_ROOT/scripts/lock-down-phase2.sh"

Enables the application firewall, stealth mode (no ICMP reply), logging; sets pmset for secure standby; disables sudden-motion-sensor; increases file/proc limits.

Verify:

/usr/libexec/ApplicationFirewall/socketfilterfw --getglobalstate     # "State = 2" = blocking
/usr/libexec/ApplicationFirewall/socketfilterfw --getstealthmode     # "on"

4. Run phase 3 — pf anchor for telemetry endpoints

bash "$DOME_ROOT/scripts/lock-down-phase3.sh"
bash "$DOME_ROOT/scripts/pf-reload.sh"

Writes /etc/pf.anchors/com.dsh with block rules for known telemetry TCP/UDP ports (5228/5229/5230 Google push, 5353 mDNS, etc.) and reloads pf.

Verify:

sudo pfctl -sa | grep com.dsh        # anchor loaded
sudo pfctl -a com.dsh -s rules       # rules active

5. Run phase 4 — daemon audit + permanent removal

bash "$DOME_ROOT/scripts/lock-down-phase4.sh"

Strictest phase: scans /Library/LaunchDaemons and ~/Library/LaunchAgents in a loop, permanently removes any plist not matching the approved allowlist (homebrew.*, com.apple.*, com.openssh.*, org.cups.*). Logs every removal to logs/daemon-watch.log.

Prompt before: some apps (e.g. Dropbox, 1Password) install LaunchAgents the user may actually want. If you see a removal that looks legitimate, pause and ask.

6. Verify the end state

Run the three audit commands and require all green:

bash "$DOME_ROOT/scripts/audit.sh"         # security posture check
bash "$DOME_ROOT/scripts/dome-check.sh"    # full protocol enforcer (includes machine-probe refresh)
"$DOME_ROOT/.venv/bin/python" -c "
import sys; sys.path.insert(0, '$DOME_ROOT')
from agents.core.machine import security_posture
import json
print(json.dumps(security_posture(), indent=2))
"

Expected security_posture() after lockdown:

{
  "filevault": true,
  "sip": true,
  "gatekeeper": true,
  "firewall": true,
  "dns_private": true,
  "secrets_backend_present": true
}

If any value is false, tell the user exactly which check and the exact fix command (see the per-phase verify steps above).

7. Schedule ongoing enforcement

The setup script should have installed a cron entry for dome-check.sh every 6h. Confirm:

crontab -l | grep dome-check

If missing, add it:

(crontab -l 2>/dev/null | grep -v dome-check; echo "0 */6 * * * cd $DOME_ROOT && bash scripts/dome-check.sh >> logs/dome-check.log 2>&1") | crontab -

8. Report

Single-block summary for the user:

✅ Sovereign Lockdown complete
✅ Firewall + stealth active
✅ DNS pinned to dnscrypt-proxy
✅ pf anchor loaded (com.dsh)
✅ N unauthorized LaunchAgent(s) removed → logs/daemon-watch.log
✅ dome-check.sh scheduled every 6h

If any step was skipped or failed, say so explicitly. No lockdown theater.

Non-negotiables

  • Never proceed past a failing phase. If phase 1 fails (dnscrypt-proxy won't start), phase 3's pf anchor will leak DNS.
  • Never remove a daemon the user depends on without asking. 1Password Agent, Dropbox, Tailscaled are legitimate.
  • FileVault cannot be auto-enabled. If security_posture()['filevault'] is False, tell the user to enable it manually in System Settings → Privacy & Security → FileVault.
  • SIP cannot be enabled from a running system. It requires a reboot into Recovery — if disabled, that's a user task, not this skill's.
  • Reversing lockdown is not this skill's job. If the user wants to undo, point them to sudo /usr/libexec/ApplicationFirewall/socketfilterfw --setglobalstate off and sudo networksetup -setdnsservers Wi-Fi empty.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

ai-sdk

無料

Vercel AI SDK expert guidance. Use when building AI-powered features — chat interfaces, text generation, structured output, tool calling, agents, MCP integration, streaming, embeddings, reranking, image generation, or working with any LLM provider.

日本語の概要は準備中です。原文の説明を表示しています。

garochee33/DSH42026年10月9日 更新

Algorithm design and analysis skill for DSH; use when working on optimization, data structures, or computational procedures.

日本語の概要は準備中です。原文の説明を表示しています。

garochee33/DSH42026年10月9日 更新

Design and implement scalable API gateways, RESTful APIs, GraphQL endpoints, WebSocket handlers, and microservice communication patterns. Use when creating API routes, designing endpoints, implementing middleware, or setting up service mesh architectures.

日本語の概要は準備中です。原文の説明を表示しています。

garochee33/DSH42026年10月9日 更新

best-of-n

無料

Implement a task N ways in parallel and pick the best. Spawns multiple subagents in isolated worktrees, evaluates all candidates, and applies the winner. Use when asked to "best of n", "try multiple approaches", "parallel implementations", "/best-of-n", or "/bon".

日本語の概要は準備中です。原文の説明を表示しています。

garochee33/DSH42026年10月9日 更新

check

無料

Check your work with a verification subagent. Spawns a verifier that reviews diffs, runs builds and tests, and evaluates correctness. Use when asked to "check work", "verify changes", "self-verify", "/check", "/verify", "/check-work", or "/self-verify".

日本語の概要は準備中です。原文の説明を表示しています。

garochee33/DSH42026年10月9日 更新

CI/CD pipeline design, GitHub Actions workflows, deployment automation, infrastructure as code, and release management. Use when setting up CI/CD, automating deployments, configuring build pipelines, or managing releases.

日本語の概要は準備中です。原文の説明を表示しています。

garochee33/DSH42026年10月9日 更新

garochee33 のスキルをすべて見る

このスキルの問題を報告する