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

# Data licensing

> License private data to Metaphi AI. It is assessed on your own machine, and nothing is sent until you agree.

Metaphi AI licenses private data to build training environments for AI agents. These pages are for data providers: the company that owns the data, or a broker acting for one.

| Data | Examples | How to start |
| - | - | - |
| Code | Git repositories with their history and tests, in Go, Python, JavaScript and TypeScript, Java, C# and Rust | [CodeVal](https://docs.metaphi.ai/kennet/data-licensing/codeval.md) |

For other kinds of data, write to [agent@metaphi.ai](mailto:agent@metaphi.ai).

## How it works

1. **Assess.** You run one command, `kennet codeval`, over the repositories you offer. It builds and tests them on your machine and sends typed measurements, after you approve the sign-in in your browser. Your source code stays with you.
2. **Decision.** Metaphi AI reviews the results. If they are accepted, Metaphi AI emails you an offer.
3. **Agreement.** Terms are put in place under a mutual NDA, which you can sign from the command line with `kennet mnda sign`.

## Start

```sh theme={"dark"}
curl -fsSL https://github.com/metaphi-labs/kennet-cli/releases/latest/download/install.sh | sh
kennet codeval ./your-repository
```

The full steps, what is sent, and every option are in [CodeVal](https://docs.metaphi.ai/kennet/data-licensing/codeval.md).

## With a coding agent

```sh theme={"dark"}
kennet agents install
```

This gives the coding agents on your machine the `kennet-partner` skill, a guide to running CodeVal and signing the mutual NDA. Start a new session and ask the agent to run CodeVal on your repositories. Nothing is sent before you approve the sign-in in your browser.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.