# Datris

> Datris is the data control plane for AI agents: an open-source, self-hosted platform that an agent drives through one MCP server to acquire data, validate it, transform it, land it in the stores a team already runs, and record every run, with provenance on landed rows when stamping is on. It sits beside the warehouse or lake; it does not replace them.

Credentials stay in the platform's vault, each agent gets its own scoped key, and an agent policy and audit log decide and record what runs. Datris runs on your own infrastructure with Docker.

## Install
- [Install for agents](https://docs.datris.ai/install-for-agents): headless install, health check, and connecting an MCP client, end to end
- Install command (needs Docker with Compose): `curl -fsSL https://get.datris.ai/install.sh | sh`

## Connect
- MCP Registry name: `io.github.datris/datris`
- [Agent skill](https://docs.datris.ai/agent-skill): teaches a coding agent when and how to use Datris in a project

## Documentation
- [Docs index for LLMs](https://docs.datris.ai/llms.txt)
- [Full docs in one file](https://docs.datris.ai/llms-full.txt)

## Source
- [GitHub repository](https://github.com/datris/datris-platform-oss)