# Exercise and finish a run

Use one meaningful bounded write, then finish or cancel without overstating empty evidence.

## Use this when

Use this after a read has confirmed connection and the run is active.

## Before you start

Know the one state change that represents the agent job. Check the [supported Admin GraphQL surface](/help/admin-api-surface.html#admin-api-surface).

## Steps

- Let the agent perform its representative write.

- Inspect whether the call was accepted, rejected, or unsupported and whether tracked state changed.

- Advance Store time only when the scenario needs a later modeled consequence.

- Finish when the intended evidence is present, or cancel safely when the run should stop.

## Success looks like

The concluded receipt distinguishes accepted calls, blocked calls, state deltas, and empty or partial evidence.

## If it fails

A zero-write simulation run does not establish impact. Cancellation preserves evidence already recorded; infrastructure start failures or runs with no agent execution do not consume a simulation run.

## Next action

Open [Read the current receipt](/help/reading-a-receipt.html#reading-a-receipt).