Meet AutoDev: Microsoft's AI-driven software development framework

Microsoft researchers revealed AutoDev, an AI-powered software development framework allowing AI agents to tackle various software development tasks autonomously, including complex code editing, advanced Git operations, and comprehensive testing.

The framework is designed to promote autonomy, efficiency, and security. AutoDev leverages Docker containers to perform housing operations to ensure streamlined and secure development processes, securing user privacy and project integrity.

This approach to security showcases Microsoft’s commitment to providing developers with a robust and reliable solution for modern software development challenges.

The primary purpose behind AutoDev is its ability to delegate complex engineering tasks to AI agents equipped with diverse tools and operations.

These AI agents efficiently navigate tasks with exceptional autonomy, seamlessly managing operations like compiling code, editing files, or executing tests.

The automation of these tasks can revolutionize the way developers work. This will allow them to focus on higher-level strategic tasks while AI agents handle the small tasks.

AutoDev Architecture

AutoDev Features

Here are the features of the framework mentioned on GitHub:

• Languages support: Java, Kotlin, JavaScript/TypeScript, Rust, Python, Golang, C/C++/OC, or others…
• Auto development mode.
  • AutoCRUD(Spring framework）. With DevTi Protocol (like devti://story/github/1102) will auto generate Model-Controller-Service-Repository code.
  • AutoSQL. Context-aware SQL generation.
  • AutoPage. Context-aware Page generation.
  • AutoTesting. create unit test intention, auto run unit test and try to fix test.
  • AutoDocument. Auto generate document.
  • AutoArkUI. Auto generate HarmonyOS ArkUI code.
• Copilot mode
  • AutoDev will help you find bugs, explain code, trace exceptions, generate commits, and more.
  • Pattern specific.Based on your code context like (Controller, Service import), AutoDev will suggest you the best code.
  • Related code. Based on recent file changes, AutoDev will call calculate similar chunk to generate the best code.
• Chat mode
  • Chat with AI.
  • Chat with selection code.
  • Chat with code context-aware (To be implemented).
• Customize.
  • Custom specification of prompt. For example, Controller, Service, Repository, Model, etc.
  • Custom intention action. You can add your own intention action.
  • Custom LLM Server. You can customize your LLM Server in Settings -> Tools -> AutoDev
  • Custom Living documentation.
  • Team prompts. Customize your team prompts in codebase, and distribute to your team.
  • Custom AI Agent. You can integrate your own AI Agent into AutoDev.
  • Prompt override. You can override AutoDev’s prompt in your codebase.
• Infrastructure / DevOps
  • CI/CD support. AutoDev will auto generate CI/CD config file.
  • Dockerfile support. AutoDev will auto generate Dockerfile.
• DevIns language
  • Executable AI Agent language.
• Built-in LLM Fine-tune
  • UnitEval evaluate llm result
  • UnitGen generate code-llm fine-tune data.

AutoDev is another sign of advancement in the field of software development. With the framework, AI’s capabilities are extended beyond simple code suggestions, which means we can expect a more efficient, robust, and secure approach while handling software engineering.

To learn more about it, you can visit AutoDev’s GitHub page. What do you think about this new framework? Share your opinions with our readers in the comments section below.