fix(ci): make health check CI-mode aware to prevent deployment timeouts

The health check was failing in CI mode because it expected sshd to be
running, but CI mode intentionally skips SSH daemon startup to use
Fly.io's hallpass service instead. This caused continuous health check
failures and deployment timeouts.

Changes:
- Health check now passes immediately in CI_MODE=true
- Production mode still validates sshd is running
- Prevents 10-minute deployment timeouts in CI workflows

Fixes deployment failures in integration-resilient workflow.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

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

Author: pacphi

docker/scripts/health-check.sh

@@ -1,7 +1,13 @@
 #!/bin/bash
 # Health check script for Docker container
 
-# Check if SSH daemon is running
+# In CI mode, we use Fly.io's hallpass SSH instead of custom sshd
+# Health check should pass as long as the container is running
+if [ "$CI_MODE" = "true" ]; then
+    exit 0
+fi
+
+# In production mode, check if SSH daemon is running
 if pgrep sshd > /dev/null; then
     exit 0
 else