CrashLabdocs
Get started

Choose your first workflow

Pick a small agent behavior that produces trustworthy value quickly.

The smallest useful evaluation protects one decision you would care about in a pull request. It should be small enough to understand, realistic enough to exercise the production agent, and strict enough to reject a plausible bad outcome.

A good first evaluation

Look for a workflow with all four of these properties:

PropertyUseful first caseAvoid initially
TriggerOne concrete user request or eventAn open-ended multi-day workflow
Side effectOne or two observable tool or state changesA broad set of loosely related actions
OutcomeA protected fact that code can inspect“The answer seems good” with no evidence boundary
FailureAt least one meaningful invariant or forbidden actionOnly checking that the process exited successfully

Examples include routing one support request without sending a reply, looking up one record before drafting an answer, or creating one ticket without changing unrelated customer data.

Write the behavioral contract first

Before implementing the world, answer these questions in plain language:

  1. What starts the workflow?
  2. What production agent entrypoint receives it?
  3. What can the agent read and change?
  4. What final state must exist for the evaluation to pass?
  5. What must never happen?
  6. Which facts must stay protected from the target?

Those answers become the evaluation's coverage description, world boundary, and verifier checks. They also let another developer decide whether the test protects the behavior they think it protects.

Prefer observable effects

Use deterministic checks for tool arguments, state changes, invariants, request budgets, and causal milestones. Add a bounded semantic judge only when the property genuinely depends on meaning that structured evidence cannot establish. Do not use a model judge to replace an exact check that code can perform.

Define reliability after the case is useful

One trial is the fastest way to debug setup. It is not a reliability claim. Once the world, target, and verifier behave correctly, run several independent trials and preserve the denominator. Start with the configured three-trial check, then increase repetitions only when the cost and decision risk justify it.

Continue with Set up an agent when you can describe one workflow at this level.

On this page