IBM wxflows
FreeIBM 官方工具平台，为任何数据源构建、测试和部署工具。
FreeFree tier
About IBM wxflows
IBM wxflows is a tool for building, testing, and deploying AI tools that integrate with watsonx.ai Flows Engine using the Model Context Protocol (MCP). The GitHub repository provides a step-by-step tutorial for setting up a wxflows project, including installing the CLI, deploying a project to a Flows Engine endpoint, and running an MCP server locally. The example project includes pre-configured integrations with Google Books and Wikipedia tools, allowing developers to quickly create AI applications that call external APIs. wxflows supports JavaScript/TypeScript via the wxflows SDK, environment variable management, and can be used with MCP clients such as Claude Desktop.
Key Features
MCP (Model Context Protocol) support for tool integration
wxflows CLI for project deployment and management
Pre-configured tool examples: Google Books and Wikipedia
wxflows SDK for JavaScript/TypeScript
Environment variable configuration for API keys
Build and run an MCP server locally
Compatible with MCP clients like Claude Desktop
Free sign-up for watsonx.ai Flows Engine account
Pros & Cons
Pros
- Free to sign up and get started
- Open-source (on GitHub)
- Easy deployment with CLI and SDK
- Follows MCP standard for tool interoperability
- Integrated with IBM watsonx.ai platform
Best For
Building AI applications that integrate external APIs via toolsDeploying and testing tools for watsonx.ai Flows EngineUsing MCP-compatible clients (e.g., Claude Desktop) to call custom tools
FAQ
How do I deploy a wxflows project?
Move into the wxflows directory of your project and run 'wxflows deploy' to deploy the endpoint and tools to the Flows Engine.
What tools are included in the example?
The example project includes Google Books (search books) and Wikipedia (search pages) tools.
Can I use wxflows with Claude Desktop?
Yes, after building the MCP server, you can configure it in Claude Desktop's configuration file to see and call the tools.