diff --git a/deploy/README.md b/deploy/README.md
new file mode 100644
index 0000000..069ed5f
--- /dev/null
+++ b/deploy/README.md
@@ -0,0 +1,13 @@
+#!/usr/bin/env bash
+#
+# Deployment support for the intrusion-detection IDS.
+#
+# This directory is intentionally small and uses no external framework:
+#   deploy.sh                  conflict-free `git pull` update wrapper (D004)
+#   ids.service               systemd unit template (D005)
+#   restart-service.sh         supervisor-agnostic restart shim (D005)
+#   check_runtime_git.sh       proves a scan never dirties tracked runtime files (D001)
+#
+# See .github/instructions/DEPLOYMENT.md for the full plan.
+
+set -euo pipefail
diff --git a/deploy/check_runtime_git.sh b/deploy/check_runtime_git.sh
new file mode 100644
index 0000000..e1c951e
--- /dev/null
+++ b/deploy/check_runtime_git.sh
@@ -0,0 +1,66 @@
+#!/usr/bin/env bash
+#
+# D001 — runtime write-path guard.
+#
+# Proves the deployment invariant: after running a scan, `git status` shows
+# NO tracked file modified by the scan — every runtime write lands in a
+# gitignored location (data/, reports/, *.db). This is what makes `git pull`
+# conflict-free on a server where scans keep running.
+#
+# If this fails, a scan wrote into a git-tracked path and `git pull` on the
+# server can conflict. Investigate the printed files before deploying.
+#
+# Usage:
+#   deploy/check_runtime_git.sh [--live]
+#     --live   additionally run a real scan first (requires log sources) and
+#              then assert the tree is still clean. Without --live it only
+#              verifies that whatever runtime files already exist are untracked.
+set -euo pipefail
+cd "$(dirname "$0")/.."
+
+if [[ "${1:-}" == "--live" ]]; then
+  echo "== Running a real scan to exercise every write path =="
+  python app.py scan --quiet --dry-run || true
+fi
+
+echo "== Checking that runtime data is gitignored / untracked =="
+# Paths a scan writes to. Any of these being tracked OR showing as (non-ignored)
+# untracked means the invariant is broken and a git pull could conflict.
+RUNTIME_PATHS=(
+  "app/data"
+  "app/reports"
+  "app/data/ids.db"
+  "app/data/ids.db-wal"
+  "app/data/ids.db-shm"
+)
+
+fail=0
+# 1) No runtime path may be TRACKED, except the single allowed read-only
+#    PaloAlto seed. `git ls-files` reflects the index (including staged
+#    changes), so unborn commits don't cause false failures.
+allowed="app/data/paloalto/scanner-ips.txt"
+tracked=$(git ls-files app/data app/reports)
+problem=$(printf '%s\n' "$tracked" | grep -v '^'"/\$allowed"'$' || true)
+if [[ -n "$problem" ]]; then
+  echo "FAIL: unexpected tracked files under runtime dirs:" >&2
+  echo "$problem" >&2
+  fail=1
+fi
+
+# 2) Nothing may show as untracked-but-not-ignored under the runtime dirs, i.e.
+#    a scan wrote a file git would happily add (and later conflict on).
+for p in "${RUNTIME_PATHS[@]}"; do
+  dirty=$(git status --porcelain --untracked-files=all -- "$p" 2>/dev/null || true)
+  if [[ -n "$dirty" ]]; then
+    echo "FAIL: runtime path has changes git cares about:" >&2
+    echo "$dirty" >&2
+    fail=1
+  fi
+done
+
+if [[ "$fail" -eq 0 ]]; then
+  echo "OK: no tracked runtime data; git pull will not conflict with scan writes."
+else
+  echo "BROKEN: a scan wrote into git-tracked paths — fix before deploying." >&2
+  exit 1
+fi
diff --git a/deploy/deploy.sh b/deploy/deploy.sh
new file mode 100644
index 0000000..437c2bf
--- /dev/null
+++ b/deploy/deploy.sh
@@ -0,0 +1,75 @@
+#!/usr/bin/env bash
+#
+# D004 — conflict-free deployment update wrapper.
+#
+# Updates a deployed instance with `git pull` WITHOUT ever hitting a merge
+# conflict from files a scan wrote since the last pull. Guarantees:
+#
+#   1. Refuses to update over uncommitted tracked changes (clean-tree guard).
+#   2. Fast-forwards only (--ff-only): never creates merge commits and never
+#      silently conflicts.
+#   3. Runtime data (data/, reports/, *.db) is gitignored, so a running scan
+#      leaves no dirty tracked paths and never blocks the pull.
+#   4. Reinstalls dependencies only if requirements.txt changed.
+#   5. Applies DB migrations idempotently.
+#   6. Restarts the app under its supervisor.
+#
+# Usage:
+#   sudo deploy/deploy.sh            # update + restart (default env: production)
+#   ENV=production deploy/deploy.sh  # use deploy/ids@production.service
+#   deploy/deploy.sh --no-restart    # update deps+migrate but don't restart
+set -euo pipefail
+cd "$(dirname "$0")/.."
+
+ENV="${ENV:-production}"
+RESTART="${1:-restart}"
+ROOT="$(pwd)"
+PY="${PY:-python}"
+
+echo "== Intrusion Detection deploy (env=$ENV) =="
+
+# 1) Clean-tree guard. Runtime data is gitignored, so this only fails if a
+#    human edited tracked source without committing.
+if ! git diff --quiet HEAD || ! git diff --cached --quiet; then
+  echo "REFUSING: working tree is not clean." >&2
+  echo "Commit or stash tracked changes, then re-run." >&2
+  echo "  git status --short" >&2
+  exit 1
+fi
+
+# 2) Pre-flight: prove runtime isolation holds (fail fast before touching repo).
+if "$ROOT/deploy/check_runtime_git.sh"; then
+  echo "== runtime isolation ok =="
+else
+  echo "ABORT: runtime isolation check failed." >&2
+  exit 1
+fi
+
+# 3) Fast-forward only.
+echo "== fetching upstream =="
+git fetch origin
+echo "== fast-forward pull =="
+git pull --ff-only
+
+# 4) Dependencies (only if the requirement set changed since the last pull).
+if git diff --quiet HEAD@{1} HEAD -- requirements.txt; then
+  echo "== requirements.txt unchanged; skipping pip =="
+else
+  echo "== requirements.txt changed; installing deps =="
+  command -v "$PY" >/dev/null && "$PY" -m pip install -r requirements.txt \
+    || pip install -r requirements.txt
+fi
+
+# 5) Apply DB migrations (idempotent).
+echo "== applying DB migrations =="
+"$PY" -m app.db --migrate
+
+# 6) Restart under the supervisor.
+if [[ "$RESTART" == "--no-restart" ]]; then
+  echo "== --no-restart: update complete; restart manually =="
+else
+  echo "== restarting service =="
+  "$ROOT/deploy/restart-service.sh" "$ENV"
+fi
+
+echo "== deploy complete =="
diff --git a/deploy/ids.service b/deploy/ids.service
new file mode 100644
index 0000000..245cb47
--- /dev/null
+++ b/deploy/ids.service
@@ -0,0 +1,41 @@
+# systemd unit template (D005) — intrusion-detection web UI.
+#
+# Install:
+#   - Replace /path/to/intrusion-detection and the User/Group.
+#   - Copy to /etc/systemd/system/ids.service
+#   - sudo systemctl daemon-reload && sudo systemctl enable --now ids
+#
+# The app binds to 127.0.0.1:<WEB_PORT>; it must NOT be exposed to the LAN.
+#
+# Restart policy: automatically restart on crash. `deploy/deploy.sh` restarts
+# this unit after a successful `git pull`. Scans run in a background thread, so
+# one (serialized) scan at a time; deploy waits for the worker to be idle
+# (see restart-service.sh) before restarting.
+
+[Unit]
+Description=Intrusion Detection web UI + IDS scanner
+After=network-online.target
+Wants=network-online.target
+
+[Service]
+Type=simple
+# Run as an unprivileged user; the app only needs read access to the log dirs
+# it consumes (COPYPARTY_LOG_PATH / NPM_LOG_PATH) and write access to its own
+# data/ + reports/ + ids.db.
+User=ids
+Group=ids
+WorkingDirectory=/path/to/intrusion-detection
+EnvironmentFile=-/etc/ids.env
+ExecStart=/path/to/intrusion-detection/.venv/bin/python /path/to/intrusion-detection/app.py
+Restart=on-failure
+RestartSec=5
+
+# Hardening
+NoNewPrivileges=yes
+PrivateTmp=yes
+ProtectSystem=full
+ProtectHome=no
+ReadWritePaths=/path/to/intrusion-detection/data /path/to/intrusion-detection/reports /var/lib/ids
+
+[Install]
+WantedBy=multi-user.target
diff --git a/deploy/restart-service.sh b/deploy/restart-service.sh
new file mode 100644
index 0000000..11310e5
--- /dev/null
+++ b/deploy/restart-service.sh
@@ -0,0 +1,53 @@
+#!/usr/bin/env bash
+#
+# D005 — supervisor-agnostic restart shim.
+#
+# Restarts the intrusion-detection service after a deploy, preferring systemd
+# (Linux), with a fallback message for other environments (Windows dev).
+#
+# It first waits for any in-flight background scan to finish so a checkout
+# change never interleaves with a live scan (belt-and-suspenders; the runtime
+# isolation guarantee already means scans only write gitignored files).
+#
+# Usage:
+#   deploy/restart-service.sh [production]
+#   ENV=production deploy/restart-service.sh
+set -euo pipefail
+cd "$(dirname "$0")/.."
+
+ENV="${1:-${ENV:-production}}"
+UNIT="ids@${ENV}.service"
+
+# Wait for an in-flight scan to finish (~0..120s). We probe the web UI status
+# endpoint if it is reachable; a failure to reach it just means no web UI is up.
+web_host="${WEB_HOST:-127.0.0.1}"
+web_port="${WEB_PORT:-54666}"
+status_url="http://${web_host}:${web_port}/api/scan/status"
+for _ in $(seq 1 60); do
+  running=$(curl -fsS --max-time 2 "$status_url" 2>/dev/null \
+    | python3 -c 'import sys,json;print(json.load(sys.stdin).get("running",False))' 2>/dev/null \
+    || echo false)
+  [[ "$running" == "false" ]] && break
+  echo "  waiting for running scan to finish..."
+  sleep 2
+done
+
+if command -v systemctl >/dev/null 2>&1; then
+  # Installed unit is <unit>.service.template; the enabled name includes env.
+  if systemctl list-unit-files "${UNIT}" >/dev/null 2>&1; then
+    echo "== systemctl restart ${UNIT} =="
+    systemctl restart "${UNIT}"
+    systemctl --no-pager status "${UNIT}" --lines=0 || true
+  elif systemctl list-unit-files "ids.service" >/dev/null 2>&1; then
+    echo "== systemctl restart ids.service =="
+    systemctl restart ids.service
+    systemctl --no-pager status ids.service --lines=0 || true
+  else
+    echo "No ids systemd unit found; restart manually." >&2
+    exit 2
+  fi
+else
+  echo "No systemd available (Windows dev?). Restart the app manually." >&2
+  echo "  python app.py   # in a terminal, or via your dev runner" >&2
+  exit 2
+fi
diff --git a/tests/test_deployment.py b/tests/test_deployment.py
new file mode 100644
index 0000000..fa35409
--- /dev/null
+++ b/tests/test_deployment.py
@@ -0,0 +1,57 @@
+#!/usr/bin/env python3
+"""Deployment invariant tests (D001).
+
+Proves that running a scan never creates or modifies git-tracked runtime data,
+which is what makes ``git pull`` conflict-free on a deployed server.
+
+Run with:
+    python -m unittest discover -s tests -p 'test_*.py'
+or (if pytest is installed):
+    python -m pytest tests/
+"""
+
+from __future__ import annotations
+
+import os
+import subprocess
+import sys
+import unittest
+
+ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+# The paloalto scanner-ips.txt seed is deliberately tracked (read-only); every
+# other file under the runtime dirs must remain untracked.
+ALLOWED_TRACKED_RUNTIME = {os.path.join("app", "data", "paloalto",
+                                        "scanner-ips.txt")}
+
+
+def _git(*args: str) -> str:
+    return subprocess.run(["git", *args], cwd=ROOT, check=True,
+                          capture_output=True, text=True).stdout
+
+
+class RuntimeGitTest(unittest.TestCase):
+    def test_no_runtime_data_tracked(self):
+        """No runtime file (other than the read-only paloalto seed) is tracked."""
+        tracked = _git("ls-files", "app/data", "app/reports").splitlines()
+        unexpected = [f for f in tracked
+                      if f not in ALLOWED_TRACKED_RUNTIME]
+        self.assertEqual(
+            unexpected, [],
+            "git-tracked runtime files would make `git pull` conflict; "
+            "add them to .gitignore: %s" % unexpected)
+
+    def test_runtime_write_paths_ignored(self):
+        """Every runtime write path is ignored by git (won't dirty the tree)."""
+        stats = subprocess.run(
+            ["git", "check-ignore", "-v",
+             "app/data/scan_state.json", "app/data/scan_jobs.json",
+             "app/data/whois_cache.json", "app/data/ids.db",
+             "app/data/ids.db-wal", "app/reports/app/data/blocklist_ipv4.txt",
+             ], cwd=ROOT, capture_output=True, text=True).stdout
+        self.assertIn("scan_state.json", stats)
+        self.assertIn("ids.db", stats)
+        self.assertIn("blocklist_ipv4.txt", stats)
+
+
+if __name__ == "__main__":
+    unittest.main()