Use cases
Performance and cost
Cut the route payload, the slow query, and the hot path doing work nobody asked for.
Slowdowns arrive a little at a time, and nobody links them to a cause until they hurt, whether it was a page that grew or a query that lost its index.
Each ticket picks one number and changes one thing, then reports the value before and after so you know what the change did.
Add caching at one boundary, with clear rules for clearing it
Hit, miss, and invalidation all covered by tests
Cut the payload on one route
Measured before and after in the pull request
Fix the slowest query on one endpoint
Query plan and timing captured before and after
Narrow a scheduled job that scans more than it needs
Timing plus a correctness test on the narrowed scope
Remove work from a hot path that only some requests need
Benchmark before and after, behaviour tests unchanged
Set a performance budget for one route and make the pipeline enforce it
Budget fails on a deliberately seeded regression
Put the agent team to work on your backlog
Connect a repository and assign SHIP your first issue. Your first missions are free.