Idempotency
Resilience and failure-handling patterns — see SYSTEM_DESIGN_ALIGNMENT.md section 9.3.
Why we need this / what value this brings
Systems fail — resilience patterns are what determine whether a failure is a brief blip or a cascading outage.
When to use this
Apply at every point the system depends on something that can fail (a network call, an external service, a retried client request).
How to use or implement this
Add the cheapest, highest-leverage pattern first (timeouts everywhere) before more complex ones (circuit breakers, sagas).
Subtopics
Empty folder — drop notes, links, and findings here as you research.