> ## Documentation Index
> Fetch the complete documentation index at: https://docs.e6data.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Connect data sources and systems

> The ways data and external systems reach AgentFlow: semantic layers, external Cube, tools and connectors.

How-to guideApplies to: All deployments

## How data enters AgentFlow

| Path | Use it for |
| - | - |
| [Semantic layer](https://docs.e6data.com/agentflow/setup/semantic-layers.md) | Governed access to tables on an e6data cluster, with business context |
| External Cube | An existing Cube model, with the `cube_load` or `cube_sql` policy and optional e6data fallback |
| Built-in tools | SQL, schema search, join discovery, data profiling, files |
| [Agent Tools](https://docs.e6data.com/agentflow/extend/custom-tools.md) | Your own systems, through Python functions |
| Connectors | Slack, GitHub, GreptimeDB and remote MCP servers |

## Turn on connectors

Configure connectors in the project's **Settings** under **Connectors**. Each connector must also be enabled for your deployment.

## Outbound systems

Workflows deliver notifications and approvals to Slack and email, and write back to systems such as ERP, MES and CMMS through tools and action steps. See [Actions and approvals](https://docs.e6data.com/agentflow/automate/actions-and-approvals.md).


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