Area: Test infrastructure · Assignee: A
What: Build a scripted, reproducible dev/test stack — k3d cluster, Prometheus, OpenCost (+ UI), and seed workloads — that comes up from a clean machine with one command and that the integration suite targets.
Why: This is the prerequisite for the Infra-priority tickets from batch 1 (Prometheus toggling, performance baselines) and for chaos testing, and it eliminates "works on my machine" variance between interns and CI.
Acceptance criteria:
One documented command brings up the full stack from scratch; one tears it down
Integration suite runs green against the stack
Provides the hooks other tickets need: toggling Prometheus, deploying ground-truth workloads, collecting logs
Documented well enough that another intern can stand it up unaided
Blocks: #4, #10, #19
Area: Test infrastructure · Assignee: A
What: Build a scripted, reproducible dev/test stack — k3d cluster, Prometheus, OpenCost (+ UI), and seed workloads — that comes up from a clean machine with one command and that the integration suite targets.
Why: This is the prerequisite for the Infra-priority tickets from batch 1 (Prometheus toggling, performance baselines) and for chaos testing, and it eliminates "works on my machine" variance between interns and CI.
Acceptance criteria:
One documented command brings up the full stack from scratch; one tears it down
Integration suite runs green against the stack
Provides the hooks other tickets need: toggling Prometheus, deploying ground-truth workloads, collecting logs
Documented well enough that another intern can stand it up unaided
Blocks: #4, #10, #19