Playbooks
Repeatable decisions. Problem, context, approach, tradeoffs.
Problem
An agent is slow, expensive, or losing the plot mid-task.
Budget the context before adding tools
Any tool-using agent with more than a handful of tools in its definition list.
Problem
A flaky downstream is causing errors and someone has opened a PR adding retry logic.
Idempotency before retries
Synchronous HTTP or RPC between services you do not own end to end.
Problem
Something is slow and the proposed fix is to put Redis in front of it.
Measure before caching
Read paths where staleness has a real cost and the access pattern is not yet known.
Problem
Two or more services write to the same table and schema changes have become terrifying.
One writer per table
Shared-database architectures mid-migration toward services.
Problem
A service must update its database and publish an event, and sometimes only one happens.
Outbox before dual writes
Any service with a transactional store and a message broker. Especially Kafka.
Problem
Database is saturating and the team is proposing a sharding project.
Read replicas before sharding
Single-region OLTP Postgres/MySQL, under ~2 TB, read-heavy (>80% reads).