SAP Cloud ALM MCP服务器

SAP Cloud ALM MCP Server (discontinued)

This project is discontinued and is no longer maintained.

Use calmcp - CALM (Cloud ALM) MCP instead — the Node/TypeScript successor to this Rust server. It reuses the knowledge of the Cloud ALM APIs gained here, but takes a different technical direction.

The documentation below is kept for reference only. Expect no further fixes, dependency updates or new API coverage in this repository.

A Model Context Protocol (MCP) server that bridges SAP Cloud ALM APIs to AI assistants like Claude. This enables natural language interaction with SAP Cloud ALM for managing features, documents, tasks, projects, test cases, and more.

Successor: calmcp - CALM (Cloud ALM) MCP

calmcp - CALM (Cloud ALM) MCP is a read-only MCP server for SAP Cloud ALM written in TypeScript. It never creates, updates or deletes data in Cloud ALM.

Where this server exposed 75+ fine-grained tools, calmcp consolidates the same API surface into four intent-based tools:

Covered services: Tasks, Projects, Features, Documents, Process Hierarchy, Test Management (manual + automated), Analytics, BSM/Status Events, Landscape, and Cross-Library (Applications, Configurations, Developments, Interfaces).

Other differences worth knowing:

• Transports: stdio locally, or Streamable HTTP remotely — this server was stdio only.
• Deployment: runs on SAP BTP Cloud Foundry with authentication via a bound BTP Destination, plus XSUAA-protected /mcp with MCP-native OAuth (interactive sign-in from Claude Desktop, Claude Code, Cursor, VS Code) or a static API key for server-to-server callers such as Microsoft Copilot Studio.
• Install: a one-click .mcpb Claude Desktop extension, a Docker image, or node dist/index.js.
• Response size: fields projection and timebox filtering keep task lists small enough for agent hosts that choke on the raw 67-attribute task payloads.

Migrating

1. Install calmcp — see its README for the .mcpb extension, stdio, HTTP and Cloud Foundry paths.
2. Move your credentials from config.json to environment variables: CALM_TENANT, CALM_REGION, CALM_CLIENT_ID, CALM_CLIENT_SECRET for OAuth2, or CALM_SANDBOX=true with CALM_API_KEY for the SAP Business Accelerator Hub sandbox. CALM_DEBUG and CALM_TIMEOUT_SECONDS replace debug and timeout_seconds.
3. Replace tool calls: the read tools below map onto calm_list / calm_get / calm_analytics; call calm_resources to discover valid resource and provider values.
4. Write operations have no equivalent. The experimental create/update/delete tools in this server are intentionally absent from calmcp, which is read-only by design.

Features

• Full SAP Cloud ALM API Coverage: Access to 9 different APIs including Features, Documents, Tasks, Projects, Test Management, Process Hierarchy, Analytics, Process Monitoring, and Logs
• Two Authentication Modes:
  • OAuth2 client credentials for production environments
  • Static API key for SAP API Business Hub sandbox testing
• 75+ MCP Tools: Comprehensive toolset for CRUD operations across all supported APIs
• OData v4 Support: Full query builder with $filter, $select, $expand, $orderby, $top, $skip
• Debug Mode: Detailed logging for troubleshooting

Prerequisites

• Rust (1.70 or later) - Install Rust
• SAP Cloud ALM Access: Either:
  • Production: Service binding credentials (client_id, client_secret, tenant, region)
  • Sandbox: API key from SAP API Business Hub

Installation

Build from Source

# Clone and navigate to the project directory
git clone https://github.com/consetto/sap-cloud-alm-odata-mcp.git
cd sap-cloud-alm-odata-mcp

# Build the release binary
cargo build --release

# The binary will be at: ./target/release/sap-cloud-alm-mcp

Configuration

Create a config.json file in the project directory (or specify a custom path with --config).

Option A: OAuth2 Mode (Production)

For connecting to your SAP Cloud ALM tenant:

{
  "tenant": "your-tenant-name",
  "region": "eu10",
  "client_id": "sb-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx!bxxxxx|calm!bxxxxx",
  "client_secret": "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx=",
  "debug": false,
  "timeout_seconds": 30
}

Fields:

Option B: Sandbox Mode (Testing)

For testing against the SAP API Business Hub sandbox:

{
  "sandbox": true,
  "api_key": "your-sap-sandbox-api-key",
  "debug": true,
  "timeout_seconds": 30
}

Fields:

Getting a Sandbox API Key:

1. Visit SAP API Business Hub
2. Log in with your SAP account
3. Navigate to any SAP Cloud ALM API (e.g., CALM_CDM_ODATA)
4. Click "Show API Key" to reveal your key

Running the Server

Standalone (for testing)

# With default config.json
./target/release/sap-cloud-alm-mcp

# With custom config file
./target/release/sap-cloud-alm-mcp --config /path/to/config.json

# With debug mode enabled
./target/release/sap-cloud-alm-mcp --debug

With Claude Desktop

Add the server to your Claude Desktop configuration:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "sap-cloud-alm": {
      "command": "/absolute/path/to/sap-cloud-alm-mcp",
      "args": ["--config", "/absolute/path/to/config.json"]
    }
  }
}

After updating the configuration, restart Claude Desktop.

With MCP Inspector (for debugging)

npx @modelcontextprotocol/inspector ./target/release/sap-cloud-alm-mcp --config config.json

Available Tools

Note: Tools marked with ⚠️ Experimental are write operations that modify data in SAP Cloud ALM. These tools require explicit user confirmation before execution.

Features API (OData)

Documents API (OData)

Tasks API (REST)

Projects API (REST)

Test Management API (OData)

Process Hierarchy API (OData)

Analytics API (OData)

Process Monitoring API (OData)

Logs API (REST)

Example Usage with Claude

Once configured with Claude Desktop, you can interact naturally:

"List all my SAP Cloud ALM projects"

"Show me the features in project XYZ"

"Create a new task in project ABC with title 'Fix login bug' and assign it to John"

"What are the open test cases for the current sprint?"

"Show me the process hierarchy for our S/4HANA implementation"

Debug Mode

Enable debug mode for troubleshooting:

1. Set "debug": true in config.json, or
2. Use the --debug command line flag

Debug output includes:

• All MCP messages (sent/received)
• API requests and responses
• Authentication flow details
• A trace file at /tmp/sap_calm_mcp_trace_{timestamp}.log

API Endpoints

The server connects to the following SAP Cloud ALM APIs:

Production URL Pattern: https://{tenant}.{region}.alm.cloud.sap Sandbox URL: https://sandbox.api.sap.com/SAPCALM

Troubleshooting

Authentication Errors

OAuth2 Mode:

• Verify your client_id and client_secret are correct
• Ensure your service binding has the required scopes
• Check that the tenant and region match your SAP Cloud ALM instance

Sandbox Mode:

• Verify your API key is valid and not expired
• Check that sandbox mode is enabled ("sandbox": true)

Connection Issues

• Ensure your network allows connections to SAP Cloud ALM endpoints
• Check if a proxy is required and configure it at the OS level
• Increase timeout_seconds for slow connections

Debug Tips

1. Enable debug mode to see detailed logs
2. Check the trace file for full request/response details
3. Use MCP Inspector to test tools interactively

License

MIT License

Contributing

This repository is discontinued and no longer accepts contributions. Please open issues and pull requests against calmcp - CALM (Cloud ALM) MCP instead.

Disclaimer

This software is provided "as is", without warranty of any kind, express or implied.

No Responsibility

The author(s) and contributor(s) of this tool assume no responsibility or liability for any damages, losses, or consequences that may result from the use or misuse of this software. This includes, but is not limited to:

• Any kind of data loss
• Any damage to systems, networks, or data
• Any legal consequences resulting from unauthorized or improper use
• Any business losses or operational disruptions
• Any security incidents or breaches