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

# Agree with Scout or Correct the Score

> Confirm Scout's score when it's right, or correct the criteria and help Scout get it right more often in future conversations.

export const Artigo = ({titulo, href, children}) => <a className="lista-artigos-item" href={href}>
    <span className="lista-artigos-texto">
      <span className="lista-artigos-nome">{titulo}</span>
      {children ? <span className="lista-artigos-desc">{children}</span> : null}
    </span>
    <svg className="lista-artigos-seta" viewBox="0 0 24 24" width="18" height="18" aria-hidden="true">
      <path d="M9 6l6 6-6 6" fill="none" stroke="currentColor" strokeWidth="2" strokeLinecap="round" strokeLinejoin="round" />
    </svg>
  </a>;

export const Lista = ({titulo, children}) => <div className="lista-artigos">
    {titulo ? <p className="lista-artigos-titulo">{titulo}</p> : null}
    <div className="lista-artigos-itens">{children}</div>
  </div>;

After reading a conversation's analysis, you decide whether Scout's score is right or needs to change. That decision becomes the real score for that service.

<Note>
  Agreeing, correcting, and dismissing an evaluation is for Administrator, Company Manager, Area Manager, and Unit Manager. The Agent sees the evaluation and the correction, but doesn't edit. Available on all plans. The Tartini interface is in Portuguese. Screen and button names appear here exactly as on screen, with the translation in parentheses.
</Note>

***

## Agree with the Evaluation

By default, the conversation stays pending until you agree with or correct the score.

Where it is: inside the **Análise do Scout** (Scout Analysis) of each conversation, in the **Avaliação da conversa** (Conversation evaluation) section.

<Steps>
  <Step title="Choose Concordo">Click the **Concordo** (I agree) card.</Step>
  <Step title="Comment if you want">Write an optional comment, up to 500 characters.</Step>
  <Step title="Confirm">Click **Confirmar avaliação** (Confirm evaluation). The conversation leaves the Pendentes list.</Step>
</Steps>

***

## Correct the Score and Criteria

By default, the correction is made per responsible party: each agent and Tino receive their own score. Older conversations, without this split, show all criteria at once.

Where it is: same section, **Discordo** (I disagree) card.

<Steps>
  <Step title="Choose Discordo">Click the **Discordo** card.</Step>
  <Step title="Choose who to correct">Click **Corrigir este responsável** (Correct this responsible party) on the agent or on Tino you want to adjust. Whoever you don't touch keeps Scout's score.</Step>
  <Step title="Correct the wrong criteria">Fill in a score from 1 to 10 only for the criteria you disagree with. The ones left blank keep Scout's score.</Step>
  <Step title="Justify each corrected criterion">Write why Scout's score is wrong, up to 200 characters per criterion.</Step>
  <Step title="Save">Click **Salvar avaliação** (Save evaluation).</Step>
</Steps>

<Note>
  There's no single score for the entire conversation: each responsible party carries their own score. It's recalculated with the same weight Scout uses: 1 for tone of voice, guidelines, compliance, resolution, and process use, and 0.5 for an agent's process adherence.
</Note>

***

## Explain the Overall Reason

In the **Por que o Scout errou** (Why Scout got it wrong) field, you can write the conversation's context and mention an agent with `@name`. This text complements each criterion's justification, it doesn't replace it.

<Note>
  In older conversations, without a split by responsible party, this field is required and asks for at least 20 characters.
</Note>

***

## See What Changes After the Correction

The score you agreed with or corrected becomes the effective score for that responsible party in the team's numbers: in Panorama, in the breakdown by criterion, and in Desempenho. These screens also show how many scores in the scope were corrected by a manager.

The agent sees the same analysis, with the corrected score and your justification.

***

## Edit or Dismiss a Saved Evaluation

An already-reviewed evaluation appears in read-only mode, with the **Editar** (Edit) and **Descartar** (Dismiss) buttons.

<Steps>
  <Step title="Reopen to edit">Click **Editar** to change the saved score, criterion, or comment.</Step>
  <Step title="Dismiss if needed">Click **Descartar** to delete your review. The conversation goes back to Pendentes.</Step>
  <Step title="Restore if needed">If you dismissed Scout's evaluation (not your review), use **Repor nas médias** (Restore to averages) to bring it back.</Step>
</Steps>

<Note>
  There are two dismiss buttons. At the top of the analysis, it removes Scout's evaluation from the averages, because the conversation wasn't a service that could be evaluated. Inside your review, it deletes only your correction and returns the conversation to Pendentes.
</Note>

***

## Frequently Asked Questions

<AccordionGroup>
  <Accordion title="Why doesn't the conversation's score change when I correct an agent?">
    Because each responsible party has their own score when the conversation was split between agents or between Tino and an agent. There's no longer a single score.
  </Accordion>

  <Accordion title="Why does a criterion show as N/A?">
    Either it doesn't apply to that responsible party, like Process Use, which is only for Tino, or the company hasn't yet registered the standard in the Cérebro.
  </Accordion>

  <Accordion title="There are two dismiss buttons. What's the difference?">
    The one in the header removes Scout's evaluation from the averages, coverage, and calibration. The one in your review deletes only your correction and returns the conversation to the Pendentes queue.
  </Accordion>
</AccordionGroup>

***

## Related Articles

<Lista>
  <Artigo titulo="Read Scout's evaluations" href="/en/agentes/scout/avaliacoes">The list of conversations and what the analysis shows.</Artigo>
  <Artigo titulo="Track team performance" href="/en/agentes/scout/desempenho-da-equipe">How the correction factors into the team's score.</Artigo>
  <Artigo titulo="The Scout Explained" href="/en/agentes/scout/o-scout-explicado">The evaluate, validate, and learn cycle.</Artigo>
  <Artigo titulo="The Cérebro Explained" href="/en/cerebro/o-cerebro-explicado">Where the tone of voice, guideline, and compliance standards live.</Artigo>
</Lista>


## Related topics

- [Tracking Your Team with Scout and Sotto](/en/primeiros-passos/guias/acompanhando-a-equipe.md)
- [Getting Started with Tartini](/en/primeiros-passos/guias/comecando-com-o-tartini.md)
- [The Scout Explained](/en/agentes/scout/o-scout-explicado.md)
