io.github.cap-js/mcp-server

Welcome to @cap-js/mcp-server

About This Project

A Model Context Protocol (MCP) server for the SAP Cloud Application Programming Model (CAP). Use it for AI-assisted development of CAP applications (agentic coding).

The server helps AI models answer questions such as:

• Which CDS services are in this project, and where are they served?
• What are the entities about and how do they relate?
• How do I add columns to a select statement in CAP Node.js?

Table of Contents

• About This Project
• Requirements
• Setup
  • Usage in VS Code
  • Usage in Claude Code
  • Usage in OpenAI Codex
  • Usage in opencode
  • CLI Usage
• Available Tools
  • search_model
  • search_docs
• Support, Feedback, Contributing
• Security / Disclosure
• Code of Conduct
• Licensing
• Acknowledgments

Requirements

See Getting Started on how to jumpstart your development and grow as you go with SAP Cloud Application Programming Model.

Setup

Configure your MCP client (Cline, opencode, Claude Code, GitHub Copilot, etc.) to start the server using the command npx -y @cap-js/mcp-server as in the following examples.

Usage in VS Code

Example for VS Code extension Cline:

{
  "mcpServers": {
    "cds-mcp": {
      "command": "npx",
      "args": ["-y", "@cap-js/mcp-server"],
      "env": {}
    }
  }
}

Example for VS Code global mcp.json:

Note: GitHub Copilot uses the mcp.json file as source for it's Agent mode.

{
  "servers": {
    "cds-mcp": {
      "command": "npx",
      "args": ["-y", "@cap-js/mcp-server"],
      "env": {},
      "type": "stdio"
    },
    "inputs": []
  }
}

See VS Code Marketplace for more agent extensions.

Usage in Claude Code

Register the server with Claude Code:

claude mcp add cds-mcp -- npx -y @cap-js/mcp-server

Usage in OpenAI Codex

Register the server with OpenAI Codex:

codex mcp add cds-mcp -- npx -y @cap-js/mcp-server

Usage in opencode

Example for opencode:

{
  "mcp": {
    "cds-mcp": {
      "type": "local",
      "command": ["npx", "-y", "@cap-js/mcp-server"],
      "enabled": true
    }
  }
}

Rules

The following rules help the LLM use the server correctly:

- You MUST search for CDS definitions, like entities, fields and services (which include HTTP endpoints) with cds-mcp, only if it fails you MAY read \*.cds files in the project.
- You MUST search for CAP docs with cds-mcp EVERY TIME you create, modify CDS models or when using APIs or the `cds` CLI from CAP. Do NOT propose, suggest or make any changes without first checking it.

Add these rules to your existing global or project-specific AGENTS.md (specifics may vary based on respective MCP client).

CLI Usage

You can also use the tools directly from the command line.

npm i -g @cap-js/mcp-server

This will provide the command cds-mcp, with which you can invoke the tools directly as follows.

# Search for CDS model definitions
cds-mcp search_model . Books entity

# Search CAP documentation
cds-mcp search_docs "how to add columns to a select statement in CAP Node.js" 1

Available Tools

[!NOTE] Tools are meant to be used by AI models and do not constitute a stable API.

The server provides these tools for CAP development:

search_model

This tool performs fuzzy searches against names of definitions from the compiled CDS model (Core Schema Notation). CDS compiles all your .cds files into a unified model representation that includes:

• All definitions and their relationships
• Annotations
• HTTP endpoints

For MCP clients, projectPath must resolve inside one of the client's advertised workspace roots. If a client does not support roots, the server process's working directory is used. Direct CLI calls treat the explicitly provided project as trusted and continue to resolve its transitive model imports outside the project directory.

Compiled models are checked for CDS source and CAP project-configuration changes on every search_model request. A successful refresh replaces the cached model; a failed refresh keeps the last valid model and is retried on the next request. The former background-refresh setting CDS_MCP_REFRESH_MS is no longer supported.

[!IMPORTANT] Compiler workers isolate CAP state between projects; they are not operating-system or filesystem sandboxes. CAP compilation can execute project configuration such as .cdsrc.js and code from project dependencies with the server process's permissions. Only compile trusted projects. Workspace-root checks protect model-source boundaries, not against malicious code inside an authorized project.

The fuzzy search algorithm matches definition names and allows for partial matches, making it easy to find entities like "Books" even when searching for "book".

search_docs

This tool uses vector embeddings to locally search through preprocessed CAP documentation, stored as embeddings. The process works as follows:

1. Query processing: Your search query is converted to an embedding vector.
2. Similarity search: The system finds documentation chunks with the highest semantic similarity to your query.

This semantic search approach enables you to find relevant documentation even when your query does not use the exact keywords found in the docs, all locally on your machine.

Support, Feedback, Contributing

This project is open to feature requests/suggestions, bug reports, and so on, via GitHub issues. Contribution and feedback are encouraged and always welcome. For more information about how to contribute, the project structure, as well as additional contribution information, see our Contribution Guidelines.

Security / Disclosure

If you find any bug that may be a security problem, please follow our instructions at in our security policy on how to report it. Please don't create GitHub issues for security-related doubts or problems.

Code of Conduct

We as members, contributors, and leaders pledge to make participation in our community a harassment-free experience for everyone. By participating in this project, you agree to abide by its Code of Conduct at all times.

Licensing

Copyright 2025 SAP SE or an SAP affiliate company and @cap-js/cds-mcp contributors. Please see our LICENSE for copyright and license information. Detailed information including third-party components and their licensing/copyright information is available via the REUSE tool.

Acknowledgments

• onnxruntime-web is used for creating embeddings locally.
• @huggingface/transformers.js is used to compare the output of the WordPiece tokenizer.
• @modelcontextprotocol/sdk provides the SDK for MCP.