# Evaluation

**Evaluation** gives teams visibility into how agents are performing across all reviews. It aggregates data from all change reviews and surfaces trends, adoption, and the impact of different agents over time.

<figure><img src="/files/1VWYBsX5EKSkDexqh3U2" alt=""><figcaption></figcaption></figure>

### Overview

* **Overview metrics**\
  At the top, you’ll find a snapshot of activity:
  * **PRs Reviewed** - the number of changes Baz has analyzed.
  * **Accepted Suggestions** - how many AI suggestions were accepted into code.
  * **Rejected Rate** - the percentage of suggestions explicitly rejected by reviewers.
* **Accepted Suggestions by Agent Type**\
  A breakdown of which categories of agents (Security Best Practices, Code Quality & Conventions, Code Correctness, etc.) are driving the most accepted suggestions.
* **Acceptance Rate Trend**\
  A time-series graph showing how acceptance rates evolve across weeks or months. This helps track adoption, improvement, and stability of Baz in your workflow.
* **Agent Table**\
  A detailed table of all agents, including:

  * **Accepted** - suggestions merged into code.
  * **Unaddressed** - suggestions left without action.
  * **Rejection Rate** - the ratio of rejected suggestions per agent.

  This makes it easy to see which agents add the most value, and where developers may need to adjust practices.

### How to use it

1. Navigate to the **Insights** tab from the sidebar and select Evaluation.
2. Use filters (Agents, User, Time) to focus on the data that matters to you.
3. Compare adoption across agent types to identify strengths and gaps.
4. Track trends to see whether Baz’s review quality is improving over time.

## FAQ

<details>

<summary>What metrics does this dashboard provide?</summary>

Metrics include: number of PRs reviewed, accepted suggestions, rejection rate, acceptance trends over time, and breakdowns by agent type.

</details>

<details>

<summary>How can I interpret “Accepted Suggestions by Agent Type”?</summary>

It shows which classes of agents (e.g. code quality, conventions, correctness) are driving the most accepted suggestions—useful to measure agent impact.

</details>

<details>

<summary>Can I filter by user, time, or agent?</summary>

Yes — you can apply filters (Agents, User, Time period) to focus on specific slices of data.

</details>


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.baz.co/insights/evaluation.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.