fix: deploy the dashboard agent one environment at a time (#4133)

The dashboard agent deploy runs a matrix leg per environment (staging,
prod). Running them in parallel deployed the same Trigger.dev project
twice at once, and one leg failed to create the background worker while
the other succeeded. `max-parallel: 1` runs the legs sequentially so a
deploy never races itself.

Also aligns the workflow's Node version to the repo standard (`22.23.1`,
per `.nvmrc`).

Verified: re-running the staging leg on its own (no parallel prod)
deployed cleanly.

.github/workflows/dashboard-agent-deploy.yml

@@ -23,6 +23,8 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       fail-fast: false
+      # One environment at a time: parallel deploys of the same project race.
+      max-parallel: 1
       matrix:
         environment: [staging, prod]
     # Per-environment reviewer gate + source of the scoped deploy PAT.
@@ -49,7 +51,7 @@ jobs:
       - name: Setup node
         uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
         with:
-          node-version: 20.20.2
+          node-version: 22.23.1
           cache: "pnpm"
 
       - name: Install + build the CLI and the agent's deps