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

# Simulation

> Test risk outcomes with realistic scenarios before touching production payments.

**Simulación** lets your team send sample transactions through the same decision pipeline used in production — without charging a card or calling your payment processor.

## Why use it

* Validate **blocklists** and **protection rules** after changes
* Train analysts on `approve`, `review`, `challenge`, and `decline` outcomes
* Populate the **Panel** and **Monitor** with data while integrating the API
* Demonstrate value to stakeholders using controlled scenarios

<Note>
  Simulations are labeled in monitoring views. They do not move real money and should not be used as regulatory evidence by themselves.
</Note>

## How to run a simulation

<Steps>
  <Step title="Open Simulación">
    Sidebar → **Simulación** (`/dashboard/simulation`).
  </Step>

  <Step title="Pick a scenario">
    Choose a preset (low risk, high amount, suspicious device, etc.) or adjust amount, email, card BIN, and IP.
  </Step>

  <Step title="Run">
    Click **Ejecutar**. Clausum returns a risk score, decision, and session identifier.
  </Step>

  <Step title="Review results">
    Check **Panel**, **Protección**, and **Monitor de transacciones** — counts and charts update for your timezone.
  </Step>
</Steps>

## What you receive

Each run returns:

| Field            | Meaning                                                                 |
| ---------------- | ----------------------------------------------------------------------- |
| `decision`       | Recommended action: approve, review, challenge, or decline              |
| `risk_score`     | 0–100 score for prioritization                                          |
| `signal_details` | Human-readable reasons (without exposing internal engine configuration) |
| `session_id`     | Reference to find the row in monitoring tools                           |

If persistence fails, the UI shows an actionable error (usually organization setup). Contact support if provisioning was never completed.

## Best practices

* Run at least one **approve** and one **decline** scenario after changing rules
* Use distinct customer emails when testing velocity rules
* Avoid double-clicking **Ejecutar** — each click creates a new assessment row (by design)
* Pair simulation with a real **sandbox API** test before production cutover

## Next steps

<CardGroup cols={2}>
  <Card title="Protection workspace" icon="shield" href="/guides/protection-workspace">
    Tune rules and blocklists.
  </Card>

  <Card title="Real-time assessment" icon="gauge-high" href="/guides/realtime-assessment">
    Wire the same flow in your backend.
  </Card>
</CardGroup>