Debug reliability and cost

Find the point where intent, execution, and observation diverge.

Questions to answer

Is the bill caused by retries, history growth, duplicated schemas, or delegation? Did a missing result cause repeated work?

Start here

Use the problem register, including a worked investigation of a costly repeating tool loop.

Read transport behavior, prompt growth, and tool schema fidelity. Compare the actual request/response transcript and tool effects before changing prompts.

Prerequisites: a small task you can describe, its permitted effects, and basic Python for executable examples. Reading the guides needs no API key.

Evidence and coverage

The existing measurements cover scripted gateway failures and framework mechanics. A general production incident cannot be diagnosed from mock pass rates alone.

Follow the concept and build path for runnable lessons and the ecosystem map for wider coverage. Return to all journeys.