#21: deploy by pulling the images CI built, not by rebuilding on the server

The first deployments died building on the target: four images, and the
webmail build alone was enough to sink it on a small arm64 host - the
build step was killed with no BuildKit error, which is what an exhausted
or overloaded machine looks like from the outside.

The images already exist. CI publishes all four to GHCR for amd64 and
arm64 on every merge, and they are the ones the suite ran against, so
rebuilding them on the deploy target is slower, heavier and less
trustworthy than pulling them.

docker-compose.yaml therefore carries no build definitions, and
docker-compose.build.yaml is the overlay that adds them for local
development and for CI - which does have to exercise the branch rather
than the published images. Both paths were run before committing.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

.github/workflows/ci.yml

@@ -200,9 +200,8 @@ jobs:
       # start the stack at all.
       - name: Cria a rede do proxy
         run: docker network create "${COOLIFY_PROXY_NETWORK:-coolify}"
-      # --build makes compose build each service locally and tag it under the
-      # same name the published image uses, so the stack comes up without ever
-      # reaching the registry.
-      - run: docker compose --env-file .env up --build --wait
-      - run: docker compose ps
-      - run: docker compose down -v
+      # The build overlay is explicit here: CI has to exercise the code in the
+      # branch, not the images already published from main.
+      - run: docker compose -f docker-compose.yaml -f docker-compose.build.yaml --env-file .env up --build --wait
+      - run: docker compose -f docker-compose.yaml -f docker-compose.build.yaml ps
+      - run: docker compose -f docker-compose.yaml -f docker-compose.build.yaml down -v