sap_analytics_cloud_mcp

This MCP server exposes the SAP Analytics Cloud REST API as a collection of MCP tools organized into 11 service areas (Content, Data Export, Data Import, Multi Actions, Calendar, Content Transport, User Management, Monitoring, Schedule & Publication, Translation, and Smart Query) plus a ping tool and guided Prompts. It authenticates to your SAC tenant via OAuth 2.0 Client Credentials, manages CSRF tokens and session cookies, and then surface all SAC REST endpoints as discrete MCP tools you can call through any MCP-compatible client. You can use this by running the Node-based server and connecting clients such as Claude Desktop, Cursor, Claude Code, or other MCP clients to invoke the individual tools (for example, fetching Stories, exporting data, or running a calendar operation).

For Claude Desktop or Cursor, configure the MCP client to point at the running server (usually via the host/port and your environment config). The server's tools are grouped by SAC surface areas; you can call specific tools like Content > Stories, Data Export > Namespaces, or Translation > XLIFF downloads. A dedicated ping tool allows quick connectivity checks, and there are Prompts to guide common workflows (explore_content, analyze_story, system_health_check). The Smart Query tool routes queries intelligently between relational and analytical SAC endpoints based on SQL patterns.

You can also use the included Agent Client workflow by configuring SAC credentials and LLM integrations in the AgentClient setup, which enables an LLM-powered interface to issue SAC REST calls through the MCP server.

Prerequisites

• Node.js v18.x or newer
• npm (comes with Node.js)
• Git
• Access to an SAP Analytics Cloud tenant and an OAuth client (as described in the README)

Install and build

# Clone the repository
git clone https://github.com/JumenEngels/sap_analytics_cloud_mcp
cd sap_analytics_cloud_mcp

# Install dependencies
npm install

# Build the server (produces build/index.js or equivalent)
npm run build

Configure environment variables (example)

• SAC_BASE_URL: https://<tenant>.<datacenter>.sapanalytics.cloud
• SAC_TOKEN_URL: https://<tenant>.authentication.<datacenter>.hana.ondemand.com/oauth/token
• SAC_CLIENT_ID: your OAuth client id
• SAC_CLIENT_SECRET: your OAuth client secret

Run the server (as Node.js MCP server)

# Start the MCP server (as configured in mcp_config.json or via your orchestrator)
npm start

Optional: build and run with Docker (if you prefer containerized deployment)

# Build docker image
docker build -t sac-mcp-server .

# Run image with environment variables wired
docker run -e SAC_BASE_URL="https://<tenant>.<datacenter>.sapanalytics.cloud" \
  -e SAC_TOKEN_URL="https://<tenant>.authentication.<datacenter>.hana.ondemand.com/oauth/token" \
  -e SAC_CLIENT_ID="<client-id>" \
  -e SAC_CLIENT_SECRET="<client-secret>" \
  sac-mcp-server

• The server reads four required environment variables at runtime: SAC_BASE_URL, SAC_TOKEN_URL, SAC_CLIENT_ID, and SAC_CLIENT_SECRET. Ensure these are set in the runtime environment for your OAuth client and SAC tenant.
• The mcp_config.json file (generated by the setup process) should be populated with your SAC credentials for Claude Desktop or other MCP clients. If you are using Claude Desktop, you can run npm run setup to generate the file and then copy its contents into your client configuration.
• Some tools modify data and may require an allowalteration=true parameter to enable write operations. Ensure your clients supply this when performing mutating actions.
• The Smart Query tool can route queries to the OData API for relational data or to the Widget Query API for analytical workloads. Use it to leverage SAC’s data capabilities efficiently.
• If you encounter connectivity issues, verify that SAC_BASE_URL, SAC_TOKEN_URL, SAC_CLIENT_ID, and SAC_CLIENT_SECRET are correct and that the OAuth client is configured for the proper tenant and data center region.