Over the past few years, I've reviewed thousands of APIs across startups, enterprises and global platforms. Almost all shipped OpenAPI documents. On paper, they should be well-defined and interoperable. In practice, most fail when consumed predictably by AI systems. They were designed for human readers, not machines that need to reason, plan and safely execute actions. When APIs are ambiguous, inconsistent or structurally unreliable, AI systems struggle or fail outright.
At the core of oRPC's design philosophy is what the team calls "powerful simplicity." Developers can define API endpoints with a syntax that feels similar to writing standard functions, while automatically gaining production features like OpenAPI specification generation, server action compatibility, and contract-first workflow support. The library supports multiple schema validators out of the box, including Zod, Valibot, and ArkType, without requiring additional configuration.
Install the package. Add mock-mcp as a dev dependency inside your project. npm install -D mock-mcp yarn add -D mock-mcp pnpm add -D mock-mcp Configure the Model Context Protocol server. For example, Claude Desktop can launch the binary through npx: Connect from your tests. Use connect to retrieve a mock client and request data for intercepted calls. Run with MCP enabled. Prompt your AI client to run the persistent test command and provide mocks through the tools.
The Disk and LocalDisk services have a new method mkdirIfNotExists to creates directories. This is useful when using the local file system to ensure that a directory exists before saving files. Otherwise, an error can be thrown when it does not. On cloud storage services such as S3, directories are created automatically when saving files so this method does nothing.
API requests undergo increasing complexity as more functionality is added to apps, requiring careful management of client-side updates to ensure consistent interaction with server-side implementations.