> For the complete documentation index, see [llms.txt](https://docs.zerve.ai/zerve-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.zerve.ai/zerve-docs/superagents/artifacts-and-outputs.md).

# Artifacts and outputs

The versioned files a superagent's team produces and builds on, and the outputs it produces for you.

A superagent's team produces two kinds of results: **artifacts**, the working files team members build on, and **outputs**, the finished results produced for you.

|                        | Artifacts                                                           | Outputs                                                        |
| ---------------------- | ------------------------------------------------------------------- | -------------------------------------------------------------- |
| **What they are**      | Versioned files, such as datasets, models, scored results or drafts | Results for you to read, such as documents, charts and reports |
| **Who uses them**      | Other team members, and you                                         | You                                                            |
| **Where to find them** | A team member's **Artifacts** tab on the Map                        | The **Outputs** tab                                            |

## Artifacts

An artifact is a file the team publishes so that others can build on it. A worker might publish a cleaned dataset, which another worker uses to train a model, which a third uses to produce a backtest.

Each artifact has a name, a type (a free-form label such as "dataset" or "model"), a description and a version history.

### Versions

Every time an artifact is updated, a new version is added: v1, v2 and so on, with the newest marked **(latest)**. Earlier versions are kept, so the superagent can compare versions to check what changed, and you can download any of them.

A version can be a single file or a folder of files. It's either produced by a team member or uploaded by you.

### Lineage

Artifacts record where they came from. When one artifact is built from another, it's marked as **derived from** it, either because the team member said so or because Zerve worked it out. That gives you a lineage you can trace, from a final result back to the data it started from.

Because of this, an artifact can't be deleted while other artifacts are derived from it.

### Working with artifacts

| To                                | Do this                                                                                                                    |
| --------------------------------- | -------------------------------------------------------------------------------------------------------------------------- |
| **See who produced what**         | On the Map, turn on the **Artifacts** layer to see how many artifacts each team member has produced                        |
| **See a team member's artifacts** | Select the team member on the Map and open its **Artifacts** tab. Its **Input** tab shows the artifacts it's working from. |
| **Upload your own**               | Upload a file as an artifact, for example a dataset or a reference document you want the team to use                       |
| **Download**                      | Download any version                                                                                                       |
| **Delete**                        | Delete an artifact or a single version, as long as nothing is derived from it                                              |

### Artifacts in batches

In [Tournaments](/zerve-docs/superagents/batches/tournament.md) and other judged batches, entrants publish their work as artifacts, and the judges compare the published artifacts.

## Outputs

Outputs are what the team produces for you to read: documents, charts, reports and other files. The **Outputs** tab on a superagent's page lists them all, with a count, for example **Outputs (12)**. You can also see the outputs of a single team member in its **Output** tab on the Map.

## Checking results

Superagents check work before presenting it, and they can require results in a particular form, such as a number within a set range. See [How superagents approach work](/zerve-docs/superagents/superagents/how-superagents-approach-work.md#they-check-the-results).


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

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

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.zerve.ai/zerve-docs/superagents/artifacts-and-outputs.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

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.