fix(ecs): preserve validated staging worker CPU

[suguanYang]

Summary

• cherry-pick the reviewed main-branch change that sets the staging worker ECS task size to 2 vCPU / 4 GiB
• preserve explicit WORKER_CONCURRENCY=10
• update the ECS deployment contract so the staging CD source matches validated revision 12

Evidence

The one-vCPU topology is rejected by the measured worker health evidence. Two fixed 2-vCPU/4-GiB Fargate Spot tasks passed the accepted production-envelope replay through 100 jobs / 60 minutes.

Evidence: https://github.com/Ontos-AI/knowhere-api-infra/issues/22#issuecomment-5302077433

Main-branch source commit: 9937fc53
Staging cherry-pick: 6e81538f

Verification

• uv run pytest deploy/ecs/test_render_task_definitions.py -q — 7 passed
• uv run --group lint ruff check deploy/ecs/test_render_task_definitions.py deploy/ecs/render_task_definitions.py — passed
• git diff --check — passed
• branch diff is identical to the reviewed main-branch diff

A repository-wide uv run pytest -q was attempted on the main branch, but collection requires optional workspace dependencies and runtime configuration absent from the isolated worktree. The complete ECS deployment contract suite passes on both branches.

Deployment note

Creating this PR does not deploy. Merging it into staging will trigger the existing staging CD workflow, register a new task-definition revision at 2 vCPU / 4 GiB, preserve the service's current desired count, and roll the staging worker tasks.