According to recent guidance, you can start working on issues, pull requests, template repositories, or ad-hoc tasks directly from GitHub in Copilot Workspace.
This all sounds promising, but as someone who's not a developer, I'm curious about how far we've come in achieving AI-only coding. Specifically, I'm interested in moving beyond "low-code" tools, often misrepresented as "no-code," towards full-fledged software development without dealing with complex code or syntax. Ideally, we should be able to describe our requirements and let AI build the software for us.
It seems we're very close to realizing this vision.
First, some basics. Some of the things Copilot Workspace is designed to help with include:
Brainstorm and Plan: Use Copilot Workspace to brainstorm ideas and plan your tasks by describing what you want in natural language
Implement Code: Generate and implement code suggestions directly within the environment
Iterate and Edit: Easily iterate on your plans and code, making edits, regenerating suggestions, and undoing changes as needed
Test and Validate: Validate your code for correctness using an integrated terminal and secure port forwarding
Collaborate: Share your workspace with your team for feedback and collaboration, with automatic versioning of context and history
Create Pull Requests: Create pull requests directly from the workspace with a single click
Work on the Go: Use the GitHub mobile app to browse issues, repos, and PRs, and open them directly within Copilot Workspace
No comments:
Post a Comment