The Microsoft Dev Proxy team has announced the release of version 0.27, which focuses on enhancing the developer experience with several new features and improvements aimed at simplifying API specification generation, improving editing suggestions, and laying the groundwork for future AI integrations.
For readers unfamiliar with it, Dev Proxy, previously named Microsoft 365 Developer Proxy, functions as a command-line tool utilized for mimicking genuine API behaviours during app testing. With Dev Proxy, developers can create resilient applications adept at managing errors and limitations.
One of the key features in this release is the ability to generate TypeSpec definitions from intercepted requests. As explained, TypeSpec is a new method for modeling APIs and generating code across different ecosystems.
This update enables Dev Proxy to quickly create TypeSpec definitions from real traffic, making API modeling easier and faster. Similar to generating OpenAPI specifications, developers can now leverage this feature to accelerate their work.
(Dev Proxy generating a TypeSpec definition for an API, Source: Microsoft 365 Dev Blog)
Another notable addition is the experimental Dev Proxy MCP server. This feature allows users to configure Dev Proxy using natural language, providing contextual assistance and helping users discover and understand Dev Proxy’s features. The integration of the Model Context Protocol (MCP) server with AI agents aims to simplify the configuration process, enabling users to work more intuitively with the platform.
With a note that this feature is currently available in an experimental version for Visual Studio Code insiders, with broader availability through npm.
The update also includes significant improvements to Dev Proxy’s local Language Model (LM) connection. The development team states the following:
We’ve refactored the local Language Model (LM) connection to make it generic. Previously, the connection logic was tightly coupled to Ollama. With this update, we’re introducing support for connecting Dev Proxy to any language model client that exposes OpenAI-compatible APIs, making it easier to connect to the AI providers you prefer.
Additionally, Dev Proxy’s JSON schemas have been enhanced with descriptions and markdown support. As reported, this change makes it easier for developers to understand and configure properties within the Dev Proxy configuration, especially when working in Visual Studio Code.
Further updates improve the intercepting of URLs and reloading of mocks when changes occur. Previous performance issue where mock files were reloaded multiple times has been addressed, noting that Dev Proxy now reloads mock files only once, providing, as stated better stability and efficiency. The update also refines the URL definition for inclusion and exclusion, improving the overall mock handling.
The Dev Proxy Toolkit extension for Visual Studio Code has also been updated, featuring new commands to create configuration files and restart Dev Proxy, along with updated code snippets that now include the latest schema and TypeSpecGeneratorPlugin configurations.
For a comprehensive list of changes in this version, developers are encouraged to refer to the official release notes.
