Riza

Riza is an MCP server that provides isolated code execution and dynamic tool creation capabilities for large language models. It connects LLM interfaces directly to the Riza API, allowing assistants to safely run user-supplied or model-generated scripts inside a dedicated sandbox environment. Software engineers, data analysts, and AI developers use Riza to evaluate arbitrary logic without risking the host system's security. By exposing an isolated interpreter, the server allows models to run one-off scripts via simple execution commands. Additionally, it turns code into reusable assets by allowing the assistant to save scripts as persistent tools using Riza's Tools API. Users can later inspect, update, list, and run these custom tools on demand. This infrastructure enables AI workflows that require dynamic programming, data transformation, or iterative script development to operate securely within sandboxed runtimes.

Category: Cloud & Infrastructure

Tags: code-execution, runtime, sandbox, webassembly

Visit Riza

How to install and configure Riza

Follow these steps to register the Riza MCP server with Claude Desktop: 1. Obtain a Riza API key from your Riza Dashboard. 2. Open your Claude Desktop configuration file (claude_desktop_config.json). 3. Add the server under the mcpServers object: json { "mcpServers": { "riza-server": { "command": "npx", "args": [ "@riza-io/riza-mcp" ], "env": { "RIZA_API_KEY": "your-api-key" } } } } 4. Replace your-api-key with your valid Riza API key. 5. Save the configuration and restart Claude Desktop.

What you can do with Riza

• Running one-off Python or JavaScript scripts directly inside an isolated code interpreter without persisting files to disk. - Authoring reusable agent tools from scratch and persisting them to the Riza Tools API for later execution. - Inspecting saved tool implementations and updating their underlying source code directly through model instructions. - Executing previously stored custom functions safely inside sandboxed runtimes to process structured data or perform computations. - Auditing available custom operations by listing all registered tools currently associated with your Riza environment.

Key facts

• https://github.com/riza-io/riza-mcp
• Cloud & Infrastructure, Developer Tools & Code Intelligence
• code-execution, runtime, sandbox, webassembly

Part of MCP Servers

How do I install Riza MCP server?

You can install and run the server using npx. Add it to your client configuration file, such as Claude Desktop config, using the command npx with the argument @riza-io/riza-mcp. You must also supply your Riza API key in the environment variables under RIZA_API_KEY. Once added, restart your MCP client.

What can Riza MCP server do?

The server allows language models to run arbitrary code inside an isolated interpreter. It also enables agents to create, list, fetch, edit, and execute reusable tools via Riza APIs. This provides safe execution for dynamic, LLM-generated code without risking host machines.

Which MCP clients work with Riza?

Riza works with any Model Context Protocol compliant client that supports local stdio processes, including Claude Desktop. You configure the server by providing the npx execution command and passing the necessary environment variables in the client configuration.

Where do I get an API key for Riza?

You can get a free Riza API key by signing up and navigating to your account dashboard on Riza's official platform. You then paste this key into your MCP client configuration under the RIZA_API_KEY environment variable.