Fix premature Connection Refused by polling for server health instead of a hardcoded 5s sleep, since Paperclip takes ~10-15s to fully start

deploy_script.sh

@@ -449,7 +449,14 @@ main() {
         docker exec -u node paperclip_app npx paperclipai onboard --yes --bind lan >/dev/null 2>&1 &
         sleep 15
         docker restart paperclip_app >/dev/null
-        sleep 5
+        log "Waiting for Paperclip to start..."
+        for i in {1..20}; do
+            if docker exec paperclip_app curl -s http://127.0.0.1:3100/api/health >/dev/null 2>&1; then
+                log "Paperclip is up and responding!"
+                break
+            fi
+            sleep 2
+        done
         bootstrap_url=$(docker exec -u node paperclip_app npx paperclipai auth bootstrap-ceo 2>/dev/null || echo "")
         
         if [[ -n "${bootstrap_url}" ]]; then