You can’t tune a retry strategy you can’t see. This is how to instrument retries so you know whether each one succeeds (saving you money) or fails permanently (burning it). Three structured metrics turn a blind spot into actionable signal.

You built an LLM judge to grade your agent. What grades the judge? An eval you never validated is a ruler you never checked against a meter stick — and a biased judge doesn’t just add noise, it moves your headline number in a consistent direction. How to meta-evaluate a judge: the labeled set, the agreement metric that isn’t accuracy, and the drift check.

The retry decision (when to give up, how long to wait) only works if you can see inside it. Idempotency keys, retry-attempt correlation, and cost accounting are the foundation. Without them, your retry logic is flying blind — you can’t tell if you’re fixing a transient failure or burning your budget on a permanent one.

‘When not to build an agent’ made the case against the loop in the abstract — quadratic cost, serial latency, an untestable failure surface. This is the concrete follow-on: three fixed pipeline shapes (linear chain, router-plus-handlers, fan-out/fan-in) that cover most of what people default to a loop for, why each one is cheaper and more testable, and the one test for when a real loop actually earns its cost.

Retry budgets cap HOW MUCH you retry; this is about WHEN to retry at all. The decision isn’t uniform: user-facing operations, background jobs, and fleet-wide cascades each have different failure costs, different retry ceilings, and different layers where the decision lives. A cost model for when to fail fast instead.

A single agent fails by getting the task wrong. A team of agents fails in ways no single agent can: correlated collapse, diffused responsibility, context fragmentation, and consensus that converges on nothing. The four failure modes that only exist once you have more than one agent — and why adding agents can lower reliability.

Prompt caching can cut your input-token bill by 90% on the reused part of a request — or do nothing at all, with no error to tell you which. It’s an exact-prefix match with a short TTL, and small, common mistakes in how agents build requests break it silently. Here’s what actually gets cached, the five ways real agents lose the discount without noticing, and how to check whether yours is.

The default retry pattern — catch the failure, append ’that didn’t work, try again,’ resend the same messages — doesn’t give the model a clean second attempt. It gives the model a context window containing its own wrong answer, which is exactly the thing most likely to make the second attempt rhyme with the first. Why retries poison the window, how to tell a poisoning retry from a safe one, and a scrub step that keeps the constraints without keeping the wrong path.

Fine-tuning example length is a design decision, not a byproduct of your data. Pad too short and you teach a distribution you’ll never see at inference; let examples sprawl and you pay quadratic training cost to memorize context you should be retrieving. How to size training sequences to the context you’ll actually serve.

Every per-step timeout is a number someone typed in without a model behind it — too short and you kill real work in flight, too long and you pay to sit idle waiting on a hang. Both mistakes are failure modes with a price tag. A small cost model finds the number that actually minimizes total cost instead of the one that felt safe.