fix: remove slow server2 preload from claude-dev deploy workflow

Drop the server2 preload stage and rely on direct docker build pulls to avoid long-running preload stream stalls in CI.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

.gitea/workflows/deploy-claude-dev-dev.yml

@@ -22,7 +22,6 @@ jobs:
       volumes:
         - /var/packages/ContainerManager/target/usr/bin/docker:/usr/bin/docker
         - /volume1/docker/claude-dev:/claude-dev-runtime
-        - /volume1/docker/nas-dashboard/ssh/vps_terminal:/runner-ssh/server2_key:ro
     steps:
       - name: Checkout exact commit
         run: |
@@ -53,16 +52,6 @@ jobs:
           echo "CLAUDE_DEV_IMAGE_TAG=${image_tag}" >> "$GITHUB_ENV"
           echo "CLAUDE_DEV_IMAGE=${image_ref}" >> "$GITHUB_ENV"
 
-      - name: Preload base image from server2
-        if: env.ROLLBACK_ONLY == '0'
-        run: |
-          set -euo pipefail
-          if ssh -i /runner-ssh/server2_key -o IdentitiesOnly=yes -o StrictHostKeyChecking=accept-new jimmyg@158.101.140.85 "podman pull docker.io/library/node:20-bookworm-slim >/dev/null && podman save docker.io/library/node:20-bookworm-slim" | docker load; then
-            echo "Preloaded node:20-bookworm-slim from server2"
-          else
-            echo "Preload from server2 failed; continuing with registry pull during docker build"
-          fi
-
       - name: Build claude-dev image
         if: env.ROLLBACK_ONLY == '0'
         run: |