> ## Documentation Index
> Fetch the complete documentation index at: https://docs.contactship.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Test in the sandbox

> Test conversations without sending WhatsApp messages to customers. Inspect replies, triggered rules and simulated tools before assigning an agent to a channel.

export const Availability = ({lang = 'es', plan, addon, permission, route, status}) => {
  const L = lang === 'en' ? {
    plan: 'Plan',
    addon: 'Add-on',
    permission: 'Permission',
    route: 'Where',
    status: 'Status',
    addonNote: 'enabled on request',
    allPlans: 'All plans',
    beta: 'Beta',
    nuevo: 'New',
    soon: 'Coming soon'
  } : {
    plan: 'Plan',
    addon: 'Add-on',
    permission: 'Permiso',
    route: 'Dónde',
    status: 'Estado',
    addonNote: 'se activa a pedido',
    allPlans: 'Todos los planes',
    beta: 'Beta',
    nuevo: 'Nuevo',
    soon: 'Próximamente'
  };
  const items = [];
  if (plan) items.push([L.plan, plan]);
  if (addon) items.push([L.addon, `${addon} · ${L.addonNote}`]);
  if (permission) items.push([L.permission, permission]);
  if (route) items.push([L.route, route]);
  if (status) items.push([L.status, L[status] || status]);
  return <div style={{
    display: 'flex',
    flexWrap: 'wrap',
    gap: '6px 22px',
    padding: '12px 16px',
    margin: '4px 0 24px',
    border: '1px solid rgba(2, 82, 255, 0.28)',
    borderLeft: '3px solid #0252ff',
    borderRadius: '8px',
    background: 'rgba(2, 82, 255, 0.05)',
    fontSize: '13.5px',
    lineHeight: '1.5'
  }}>
      {items.map(([k, v]) => <div key={k} style={{
    display: 'flex',
    gap: '6px',
    alignItems: 'baseline'
  }}>
          <span style={{
    fontSize: '10.5px',
    fontWeight: 600,
    letterSpacing: '0.07em',
    textTransform: 'uppercase',
    opacity: 0.65
  }}>{k}</span>
          <span style={{
    fontWeight: 500
  }}>{v}</span>
        </div>)}
    </div>;
};

<Availability lang="en" permission="agents.read" route="Agents → Text → Test agent" status="nuevo" />

The sandbox lets you write as a customer and inspect the agent's reply. Tools appear as simulations with their arguments. It tests conversational behavior; it does not verify WhatsApp delivery or live connections to your systems.

## Start a test

<Steps>
  <Step title="Save your configuration">
    Save every tab you changed. Tests use the saved agent; they do not send unsaved editor content.
  </Step>

  <Step title="Open Test agent">
    Use the header button. The panel identifies the environment as **Sandbox**.
  </Step>

  <Step title="Write as a customer">
    Send a specific message, wait for the reply and continue. Also test incomplete information and requests for human support.
  </Step>

  <Step title="Inspect tools">
    Expand the simulated tool indicator to see names and arguments. The blocked-effect badge means the real action was blocked during the test.
  </Step>

  <Step title="Reset for another scenario">
    Use **Reset** to start without the previous conversation's context. Save agent changes before testing again.
  </Step>
</Steps>

## Read the panel

| Element             | Meaning                                                         |
| ------------------- | --------------------------------------------------------------- |
| Agent reply         | Text generated for this scenario                                |
| Triggered rule      | A rule produced the response; its name appears when applicable  |
| Simulated tools     | Actions the agent would have attempted, including arguments     |
| Real effect blocked | The test blocked an action with real effects                    |
| Tokens              | Usage returned with the response when available; not an invoice |
| Thinking…           | The request is in progress; the panel shows elapsed time        |

## Test scenarios

| Test message                            | Expected result according to your configuration          |
| --------------------------------------- | -------------------------------------------------------- |
| A question covered by documents         | An answer supported by available information             |
| An order request without its reference  | Asking for the missing information before querying       |
| A catalog request                       | The configured image or PDF tool                         |
| “I want to speak to a person”           | Delegation to the intended destination                   |
| A question outside the business's scope | Acknowledging the limitation without inventing an answer |
| A correction to contact information     | Correct arguments for enabled fields                     |

Record the message, expected result and observed result. Repeat these scenarios after changing instructions, model or tools.

## Test scope

The sandbox does not send customer messages through WhatsApp. A simulated tool does not confirm a CRM update, a real conversation transfer or an API response.

The HTTP tool's test request is different: it contacts the endpoint. Read [Tools](/en/text-agents/tools) before using it.

## Common problems

| Message or symptom          | What to do                                                                                         |
| --------------------------- | -------------------------------------------------------------------------------------------------- |
| Sandbox could not start     | Check that the agent exists in the current organization. Reset and keep error details for support. |
| Message could not be sent   | Check the connection and error details before repeating the test.                                  |
| An expected tool is missing | Save **Tools** and review the description and instructions that should activate it.                |

## Frequently asked questions

<AccordionGroup>
  <Accordion title="Does this test contact a customer?">
    It does not send WhatsApp messages to customers. The panel uses a sandbox chat and displays simulated tools.
  </Accordion>

  <Accordion title="Can I test unsaved changes?">
    Save the tabs you changed first. The test uses the agent identifier and does not send the editor draft.
  </Accordion>

  <Accordion title="Does a successful test mean WhatsApp is ready?">
    You still need a connected channel and a rule assigning conversations to the agent. See [Channel settings](/en/messages/channel-settings).
  </Accordion>

  <Accordion title="How do I avoid carrying over previous replies?">
    Use **Reset** before the next scenario. This prevents previous conversation context from affecting the new test.
  </Accordion>
</AccordionGroup>

## See also

* [Instructions](/en/text-agents/instructions)
* [Tools](/en/text-agents/tools)
* [Your first WhatsApp agent](/en/get-started/first-whatsapp)