Wwise-MCP is a Model Context Protocol server that enables LLMs to interact with the Wwise Authoring application through a Python-based WAAPI bridge and a set of helper tools. It exposes a suite of WAAPI-enabled capabilities, including session connection to an active Wwise project, hierarchy indexing, object creation and organization, event authoring, game object management, batch RTPC/switch/state setup, audio import and discovery, soundbank configuration and build, runtime control of audio playback, and various layout and property utilities. You can refer to the included Tools documentation to see each capability and example prompts. To begin, start Wwise, ensure WAAPI is accessible, then run the MCP server so your MCP client (Claude, Cursor, or another compatible platform) can connect and issue commands against the current Wwise session.

Typical workflow:

• Start Wwise and open a project.
• Run the Wwise-MCP server so the MCP client can connect to the active WAAPI session.
• Use prompts like “Connect to my Wwise project” and “Resolve parent path” to build an index of your project structure for fast navigation. Then issue batch operations such as creating objects, authoring events, or configuring RTPCs and states through the provided tool prompts. The server translates your natural language prompts into WAAPI calls and orchestrates actions within the Wwise project.

The tools support operations such as session connection, hierarchy indexing, batch object creation and movement, event generation, runtime control (play/stop, object positioning), parameter setup (RTPCs, switches, states), audio import, and soundbank preparation. For best results, use prompts aligned with the documented tool names and follow the recommended workflows in the Tools documentation linked in the repo.

Prerequisites:

• Python 3.13+
• Git
• A local Wwise installation (Wwise 2024.1 or newer) with WAAPI access enabled

Installation steps:

1. Clone the repository: git clone https://github.com/BilkentAudio/Wwise-MCP.git cd Wwise-MCP

2. Set up a Python virtual environment and activate it:

   macOS/Linux

   python3 -m venv venv source venv/bin/activate

   Windows

   python -m venv venv .\venv\Scripts\activate

3. Install Python dependencies for the app (if a requirements file is provided, otherwise install as needed): pip install -r app/requirements.txt

4. Verify the Python entry point exists. The main script is the wwise_mcp entry point. You can install or run directly as a module: python -m wwise_mcp or if running from source, execute the module via: python app/wwise_mcp.py

5. Start the MCP server and configure your MCP client to connect to the server using the mcp_config described in this document. Ensure Wwise is running and WAAPI is accessible from your environment.

6. Optional: follow project-specific setup docs for Claude or Cursor to integrate the Wwise-MCP server into your preferred MCP client.

Tips and caveats:

• This server is experimental. Expect breaking changes or incomplete features; verify compatibility with your Wwise project before using in production.
• Ensure your Wwise version (2024.1+) supports WAAPI and that WAAPI is accessible from the machine running the MCP server.
• macOS users may encounter Gatekeeper or security prompts; follow the repository’s macOS security guidance to allow and run the binary or script.
• If you encounter environment-related issues, use a dedicated Python virtual environment and ensure the correct Python version (3.13+) is used.
• The server is designed to work with an MCP client (e.g., Claude or Cursor). Each client may have its own JSON configuration for mcpServers; this server expects a Python-based execution entry point as described in mcp_config.