Your data model, inside every AI agent.

Each DataJoinery environment runs its own Model Context Protocol server. Connect Claude or any MCP client, sign in with OAuth 2.1, and ask about your model, lineage and pipelines in plain language.

Server URL

https://[your-environment].mcp.datajoinery.io/mcp

Book a demo See the tools

DataJoinery MCP

Signed in with OAuth 2.1

Connected

What SQL builds the customer_h hub?

get_entity_sqlcustomer_h

Here’s the generated, insert-only SQL task for customer_h. It loads new business keys with their hash key and load datetime.

Switch to the latest project version.

get_latest_versionthenchange_version

Done. You’re now working in the latest version of the project.

Connected in minutes, secured with OAuth 2.1.

Agents act with your DataJoinery sign-in. Access tokens are short-lived, and a long-lived refresh token renews them, so every request stays authenticated.

  1. Add the server

    Add your environment’s MCP server URL to Claude or another MCP client.

  2. Sign in

    The client sends you to DataJoinery to authenticate.

  3. Give consent

    Approve the agent’s access on the DataJoinery consent screen.

  4. Receive tokens

    The MCP server issues OAuth 2.1 access and refresh tokens.

  5. Ask away

    The agent makes authenticated requests, renewing its access as needed.

One server, the whole model.

The DataJoinery MCP server gives agents tools for every layer, from source systems to presentation, plus the pipelines that connect them.

Conceptual model

Entities, relationships, bridges and presentations.

search_conceptual_entitiessearch_conceptual_relationshipssearch_conceptual_bridgessearch_presentations

Facts and dimensions

The reporting layer your BI tools and agents query.

search_factsget_factsearch_dimensionsget_dimension

Entities and generated SQL

Logical and physical entities, their fields and the SQL behind them.

search_entitiesget_entityget_physical_entitiesget_entity_sql

Lineage

Summary and field-level lineage, upstream and down.

get_lineage_summary_for_logical_entityget_lineage_detailed_for_logical_entityget_logical_entities_by_conceptual_entity

Pipelines

Runs, tasks, their SQL statements and error messages.

get_pipeline_runsget_pipeline_run_tasksget_pipeline_run_task

Sources, projects and versions

Browse source databases and tables, and switch project versions.

get_source_system_databasessearch_source_system_entitiesget_projectsget_versionschange_version

Ask in plain language.

A few of the questions teams hand to their agents, and the tools that answer them.

“Did any production pipelines fail today?”

get_pipeline_runs

“Show the field-level lineage for order_l.”

get_lineage_detailed_for_logical_entity

“Why did last night’s order load fail?”

get_pipeline_run_task

“List the facts and dimensions in this project.”

search_facts, search_dimensions

“Which source tables aren’t mapped yet?”

search_source_system_entities

“What SQL builds the customer_h hub?”

get_entity_sql

Give your agents the full picture.

Speak to a DataJoinery data specialist about connecting MCP to your environment.

Book a demo Email support