diff --git a/AGENTS.md b/AGENTS.md
index 3498668..53d17cc 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -85,7 +85,7 @@ graph TD
 - `sudo` on both boxes needs a tty. The remote sudo steps in `deploy.sh` fail non-interactively. Use `echo '<password>' | sudo -S <command>`.
 - Never run `pkill -f` with a pattern that appears in your own command line. It kills your ssh session. Use `pkill -x <exact-name>`.
 - Model load is slow: about 3 minutes warm, more than 8 minutes cold. Health returns `503 Loading model` the whole time. That is normal. Do not kill the process.
-- Page-cache state confounds crash tests. Cold-cache loads are slow and survive. Warm-cache loads crash fast. Control for it when testing.
+- Page-cache state confounds crash tests. Cold-cache loads are slow and survive. Warm-cache loads crash fast. Control for it when testing. `deploy.sh` drops caches on Dankclaw before restart (see `DROP_CACHES`); manual restarts must run `sync && echo 3 | sudo tee /proc/sys/vm/drop_caches` first.
 
 ### Dankclaw memory model
 
diff --git a/docs/deployment.md b/docs/deployment.md
index aee9795..437c4f8 100644
--- a/docs/deployment.md
+++ b/docs/deployment.md
@@ -59,18 +59,22 @@ The script does these actions:
 2. Pushes the `main` branch to origin.
 3. Pulls the repository on the target machine.
 4. Installs or updates the systemd unit.
-5. Reloads systemd and restarts the service.
-6. Waits for the health endpoint to respond.
+5. Drops page caches on Dankclaw before the restart. This prevents the load stall that happens when the model's pages are fully cached.
+6. Reloads systemd and restarts the service.
+7. Waits for the health endpoint to respond (over ssh).
 
-## Manual Update
+Set `FRANK_SUDO_PW` when running headless, so that remote sudo reads the password from stdin:
 
-If you cannot run the deploy script, update a server manually. Complete each step on the target machine.
+```
+FRANK_SUDO_PW=... ./scripts/deploy.sh dankclaw
+```
 
 ### Update Dankclaw
 
 1. Pull the repository.
 2. Symlink the unit file, if it changed.
-3. Reload and restart the service.
+3. Drop the page caches.
+4. Reload and restart the service.
 
 Run these commands on Dankclaw:
 
@@ -78,6 +82,7 @@ Run these commands on Dankclaw:
 cd /home/dank/vllmScripts
 git pull origin main
 sudo ln -sf /home/dank/vllmScripts/scripts/frank-default-large.service /etc/systemd/system/frank-default-large.service
+sync && echo 3 | sudo tee /proc/sys/vm/drop_caches
 sudo systemctl daemon-reload
 sudo systemctl restart frank-default-large
 curl -sf http://127.0.0.1:7770/health
diff --git a/scripts/deploy.sh b/scripts/deploy.sh
index 31ec1b6..5361d12 100755
--- a/scripts/deploy.sh
+++ b/scripts/deploy.sh
@@ -5,14 +5,18 @@
 # AGENTS.md. It assumes the local changes are committed and pushed, then:
 #   1. Pulls the repo on the target
 #   2. Installs / enables the systemd unit
-#   3. Reloads systemd and restarts the service
+#   3. Drops page caches (dankclaw) and restarts the service
 #   4. Waits for the health endpoint and reports status
 #
+# Sudo: when FRANK_SUDO_PW is set, remote sudo pipes it via `sudo -S` for
+# headless runs. Otherwise plain `sudo` is used (tty or passwordless config).
+#
 # Usage:
 #   ./deploy.sh <target> [--dry-run]
+#   FRANK_SUDO_PW=... ./deploy.sh <target>
 #
 # Targets:
-#   dankclaw   - vLLM-ROCm server (frank-default-large.service, port 7770)
+#   dankclaw   - llama.cpp server (frank-default-large.service, port 7770)
 #   dankserver - llama.cpp server (frank-default-small.service, port 7770)
 #
 # Example:
@@ -23,19 +27,23 @@ set -euo pipefail
 # ---------------------------------------------------------------------------
 # Target configuration
 # ---------------------------------------------------------------------------
-declare -A REMOTE_USER REMOTE_HOST REMOTE_DIR SERVICE SERVICE_HEALTH
+declare -A REMOTE_USER REMOTE_HOST REMOTE_DIR SERVICE SERVICE_HEALTH DROP_CACHES
 
 REMOTE_USER[dankclaw]=dank
 REMOTE_HOST[dankclaw]=dankclaw
 REMOTE_DIR[dankclaw]="/home/dank/vllmScripts"
 SERVICE[dankclaw]="frank-default-large"
 SERVICE_HEALTH[dankclaw]="http://127.0.0.1:7770/health"
+# Drop page caches before restart: with the model's pages fully cached, the
+# KV/graph allocation stalls at load (VRAM stuck at 28 GiB). See AGENTS.md.
+DROP_CACHES[dankclaw]=1
 
 REMOTE_USER[dankserver]=dank
 REMOTE_HOST[dankserver]=dankserver
 REMOTE_DIR[dankserver]="/home/dank/vLLM"
 SERVICE[dankserver]="frank-default-small"
 SERVICE_HEALTH[dankserver]="http://127.0.0.1:7770/health"
+DROP_CACHES[dankserver]=0
 
 TARGET="${1:-}"
 DRY_RUN=0
@@ -55,6 +63,16 @@ if [ -z "${REMOTE_HOST[$TARGET]:-}" ]; then
     exit 1
 fi
 
+# ---------------------------------------------------------------------------
+# Sudo wrapper (headless-friendly)
+# ---------------------------------------------------------------------------
+# When FRANK_SUDO_PW is set, remote sudo reads the password from stdin via
+# `sudo -S`. Otherwise plain `sudo` is used (tty or passwordless config).
+SUDO_PREFIX="sudo"
+if [ -n "${FRANK_SUDO_PW:-}" ]; then
+    SUDO_PREFIX="printf '%s\n' '$FRANK_SUDO_PW' | sudo -S"
+fi
+
 # ---------------------------------------------------------------------------
 # Local checks (pre-flight)
 # ---------------------------------------------------------------------------
@@ -93,6 +111,14 @@ REMOTE="$(
 # ---------------------------------------------------------------------------
 # Remote: pull + install + restart
 # ---------------------------------------------------------------------------
+DROP_CACHES_CMD=""
+if [ "${DROP_CACHES[$TARGET]}" = "1" ]; then
+    DROP_CACHES_CMD="
+echo 'Dropping page caches for reliable model load...'
+${SUDO_PREFIX} sh -c 'sync; echo 3 > /proc/sys/vm/drop_caches'
+"
+fi
+
 REMOTE_SCRIPT="
 set -euo pipefail
 cd '${REMOTE_DIR[$TARGET]}'
@@ -103,24 +129,27 @@ git pull origin main
 # Install / symlink the systemd unit
 case '${TARGET}' in
   dankclaw)
-    sudo ln -sf '${REMOTE_DIR[$TARGET]}/scripts/frank-default-large.service' /etc/systemd/system/frank-default-large.service
+    ${SUDO_PREFIX} ln -sf '${REMOTE_DIR[$TARGET]}/scripts/frank-default-large.service' /etc/systemd/system/frank-default-large.service
     ;;
   dankserver)
-    sudo ln -sf '${REMOTE_DIR[$TARGET]}/scripts/frank-default-small.service' /etc/systemd/system/frank-default-small.service
+    ${SUDO_PREFIX} ln -sf '${REMOTE_DIR[$TARGET]}/scripts/frank-default-small.service' /etc/systemd/system/frank-default-small.service
     ;;
 esac
-
+${DROP_CACHES_CMD}
 echo '[3/4] Reload and restart ${SERVICE[$TARGET]}...'
-sudo systemctl daemon-reload
-sudo systemctl enable '${SERVICE[$TARGET]}'
-sudo systemctl restart '${SERVICE[$TARGET]}'
+${SUDO_PREFIX} systemctl daemon-reload
+${SUDO_PREFIX} systemctl enable '${SERVICE[$TARGET]}'
+${SUDO_PREFIX} systemctl restart '${SERVICE[$TARGET]}'
 "
 
 if [ "$DRY_RUN" = "1" ]; then
     echo "[dry-run] Remote commands would run:"
     echo "$REMOTE_SCRIPT"
 else
-    ssh "$REMOTE" "bash -c '$REMOTE_SCRIPT'"
+    # shellcheck disable=SC2087
+    ssh "$REMOTE" bash -s <<REMOTE_EOF
+${REMOTE_SCRIPT}
+REMOTE_EOF
 fi
 
 # ---------------------------------------------------------------------------
@@ -129,10 +158,10 @@ fi
 if [ "$DRY_RUN" = "1" ]; then
     echo "[dry-run] Would wait for health at ${SERVICE_HEALTH[$TARGET]}"
 else
-    echo "[4/4] Waiting for health endpoint ${SERVICE_HEALTH[$TARGET]}..."
+    echo "[4/4] Waiting for health endpoint ${SERVICE_HEALTH[$TARGET]} (via ssh)..."
     HEALTH_URL="${SERVICE_HEALTH[$TARGET]}"
     for i in $(seq 1 300); do
-        if curl -sf "$HEALTH_URL" > /dev/null 2>&1; then
+        if ssh "$REMOTE" "curl -sf '$HEALTH_URL'" > /dev/null 2>&1; then
             echo "  Service healthy after approx $((i * 2))s"
             echo ""
             echo "=== Deploy complete: ${TARGET} ==="