app show has no .run.stages (README triage recipe fails); no container logs #30

1. The README failure-triage recipe returns nothing

The README suggests:

eds wf app show "$APP_ID" --json | jq -r '.run.stages[].jobs[] | select(.status=="failed") | .id' \
  | xargs -I{} eds wf job logs {}

But .run in eds wf app show --json has only branch, created_at, id, pipeline_id, status, temporal_run_id, temporal_workflow_id, updated_at: there is no stages field, so jq fails. Stages and jobs are available from eds wf run show <run_id> (.stages[].jobs[]) or eds wf job list --run-id <run_id>. Either return stages in app show or update the recipe.

2. No logs of the running container

eds wf job logs covers pipeline jobs only. Logs of the running application (the container's stdout/stderr) are not available from eds at all, so after a successful deploy there is no CLI way to see why the app returns errors. Something like eds wf app logs <id> [--since …] [--follow] would close the loop.

3. eds wf job logs details (related: #2, #6)

• Newest entries come first and the default page is 10 entries, so eds wf job logs <id> shows the end of the log in reverse order.
• --limit 5000 → Error: workflow client logs: 400 Bad Request; the maximum page size is not documented (--limit 1000 works).

Environment: eds main-868af1c, macOS.