AI & Analytics Legends The knowledge platform for SAP Analytics Sign inSign up
Guide

SAP Business Data Cloud MCP: what exists for SAP BDC

SAP Business Data Cloud MCP: what exists for SAP BDC — Analytics Legends section illustration for SAP analytics market intelligence (Datasphere, BDC, SAC, BW/4HANA)

As of 2026-10-18

Edited by , sole editor and publication director — 27 years in the SAP ecosystem. How these pages are produced, and where automation is used.

SAP BDC has no official MCP server in our verified catalogue (18 October 2026). Two community servers are relevant: SAP Business Data Cloud MCP, with seven Delta Sharing tools, and SAP Datasphere MCP, for metadata, queries and lineage. Both are MIT-licensed and unsupported by SAP, and they do very different jobs.

The BDC server: Delta Sharing, not analytics

The community SAP Business Data Cloud MCP exposes seven tools for Delta Sharing in BDC: publishing data products, validating shares and provisioning shares end to end. It runs over stdio, so it lives on the machine of whoever runs the assistant.

Note what that is. It is a data-product provisioning tool, not a question-answering tool. Provisioning a share is an action that changes who can read what, which makes it a governance event. Do not let a conversational assistant do it without an approval step.

The Datasphere server: exploration

For the analytical side there is a community SAP Datasphere MCP, listed as in active development (2.x) with both transports. It lets an assistant explore Datasphere metadata, run analytic queries, extract data and inspect lineage, with configurable PII masking. Since BDC builds on Datasphere, this is the closer fit for questions such as "which views feed this data product".

Local versus remote

The BDC server is stdio only, so there is no network endpoint to defend, but the credentials sit on a workstation. The Datasphere server supports both transports; if you host it remotely, you take on authentication, TLS, rate limiting and logging yourself. Streamable HTTP is the transport for multi-tenant deployments, stdio for local agents, as our MCP mechanics notes put it.

What to check before production

Scope: separate read tools from the provisioning tools and expose only what the use case needs. Identity: confirm whether calls run as the person or as a shared technical user, because BDC data products are governed by who may share and consume them. Masking: if you rely on the Datasphere server's PII masking, test it against your own columns rather than trusting defaults. Audit: make sure the tool call, the Delta Sharing action and the Datasphere query are each traceable.

Supply chain matters too. These are individual-maintainer community projects; the catalogue shows no SAP support statement. Assign an owner, pin versions and plan how you would remove it.

A pragmatic first step is to run the Datasphere server against a non-production space with masking on and with queries only, list its tools and remove anything that provisions or deletes. Then register the server in your inventory: our notes recommend SAP AI Agent Hub as the inventory of agents, LLMs and MCP servers, and advise versioning tool contracts so that a change in a community release does not silently change what the assistant can do.

How it relates to Joule and agents

SAP's own agentic route for BDC is Joule agents inside the product, and a Datasphere data product can act as a grounding source for them. MCP is the open alternative when you want an external agent such as a coding assistant or a custom agent to reach BDC alongside other tools. Where an official route is needed, the governed path is SAP Integration Suite's MCP gateway, which can front SAP and non-SAP APIs and external MCP servers; the catalogue gives no BDC-specific artefact there.

For market-side questions about BDC, including which consultancies, rates and opportunities exist, the Analytics Legends MCP serves read-only tools, and does not touch your BDC tenant.

The catalogue does not cover Databricks-side or Fabric-side MCP servers, so nothing here should be read as describing how BDC Connect partners expose their own data. If your architecture spans SAP and a partner platform, assess each side's server separately and decide where the identity boundary sits.

What we cannot assert

Status of the BDC community server, and whether its tools can be run read-only, are not stated in the catalogue.

Frequently asked

Is there an official SAP BDC MCP server?

Not in our verified catalogue of 18 October 2026. The SAP Help Portal search found no official Datasphere or Business Data Cloud MCP page.

What does the SAP BDC MCP community server do?

It offers seven tools for Delta Sharing: publishing data products, validating shares and provisioning shares end to end, over stdio.

Can I query BDC data through MCP?

The community Datasphere server can run analytic queries and extract data, with configurable PII masking. The BDC server itself focuses on sharing.

Should an assistant be allowed to provision shares?

Treat it as a change to data governance and require human approval; do not leave it open to free-form chat.

Engage the editor

This guide is written by . 27 years of SAP — Datasphere, Business Data Cloud, SAC Planning, Joule. Freelance or permanent · Hybrid / remote · Anywhere in EMEA. Available from 19 October 2026.

Profile and contact

What this page is built on

External sources

Read next

Guides that answer with this page

These guides cite this page as one of the sources their answer rests on.