Skip to lesson
Exit
Metric Design & Product Sense1 / 3

2 min lesson

Composing SLOs across the harness

Put yourself in this case: "Four harness stages each run at 99% reliability. Why can the end-to-end agent-run reliability still feel unacceptable and what does that imply for SLO design?" Give the clearest next step.

Step 1 of 3

Composing SLOs across the harnessthe chain is only as reliable as its weakest stage

An agent run is a pipeline: retrieval, then model, then tool-call, then apply. End-to-end reliability is the product of the stage reliabilities, not the average - and the gap between that product and the average is where users feel the product fail.

RELIABILITY COMPOSES, IT DOES NOT AVERAGE

Interactive diagram. Tab through its regions; each focused region shows its detail in the panel below.

diagram: stack

End-to-end success rests on every stage below it - multiply, never average.