fix(ci): stop & remove old container before deploying new one

docker compose up --force-recreate only works when Compose owns the
container. If the container was started outside Compose (e.g. manually
via docker restart), Compose can't recreate it and errors with
"container name already in use". Explicitly stopping and removing it
first handles both cases cleanly.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

.gitea/workflows/ci-cd.yml

@@ -91,7 +91,10 @@ jobs:
           DOCKER_BUILDKIT: 1
 
       - name: Deploy
-        run: docker compose up -d --no-deps --force-recreate api
+        run: |
+          docker stop drsousan_api 2>/dev/null || true
+          docker rm   drsousan_api 2>/dev/null || true
+          docker compose up -d --no-deps api
 
       - name: Wait for healthy
         run: |