Frequently Asked Questions about Furikake and MCP Servers

Q: What is Furikake (furi)?

A: Furikake is a local CLI & API designed for easy management of MCP (Model Context Protocol) servers. It allows you to download, manage, and run MCPs directly from your command line or through HTTP API routes.

Q: What is an MCP Server?

A: An MCP (Model Context Protocol) server acts as a bridge, allowing AI models to access and interact with external data sources and tools. It standardizes how applications provide context to Large Language Models (LLMs).

Q: What are the key benefits of using Furikake?

A: Furikake simplifies MCP management with features like easy MCP downloading from GitHub, a fully featured CLI, HTTP API routes, process state management, log management, and integration with npm.

Q: How do I install Furikake?

A: First, install Bun using the command curl -fsSL https://bun.sh/install | bash. Then, you can use Furikake’s CLI commands to manage MCPs.

Q: What operating systems does Furikake support?

A: Furikake supports macOS and Linux operating systems.

Q: How do I add an MCP using Furikake?

A: Use the command furi add <author/repo>. For example, furi add smithery-ai/mcp-fetch.

Q: How do I start an MCP using Furikake?

A: Use the command furi start <author/repo> -e <env>. The -e <env> option is for specifying environment variables, which depends on the MCP.

Q: How do I access an MCP via HTTP API?

A: First, turn on the HTTP proxy with furi http on. Then, you can access various routes like /api/status, /api/tools, and specific MCP routes like /api/<mcp-name>/status.

Q: What is UBOS and how does Furikake relate to it?

A: UBOS is a full-stack AI Agent Development Platform. Furikake is a tool that enhances the UBOS ecosystem by simplifying the management of MCP servers, which are crucial for connecting AI Agents with data and tools.

Q: Can I contribute to Furikake?

A: Yes! You can open an issue or a pull request on the Furikake repository. Contributions are welcome.