fix(ci): fix health check timeout and per-profile MCP server URL routing

The Nextcloud health check expected HTTP 401 from serverinfo, but NC 32
returns 200 — causing 5-minute timeouts. Switch to /status.php with
"installed":true check (matches Docker healthcheck). Also route the
correct MCP_SERVER_URL per CI matrix profile into the app container so
Astrolabe connects to the right service, and make the init script
gracefully skip when the var is unset.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

docker-compose.yml

@@ -47,6 +47,7 @@ services:
       - MYSQL_USER=nextcloud
       - MYSQL_HOST=db
       - REDIS_HOST=redis
+      - MCP_SERVER_URL=${MCP_SERVER_URL:-}
     healthcheck:
       test: ["CMD-SHELL", "curl -Ss http://localhost/status.php | grep '\"installed\":true' || exit 1"]
       interval: 10s