Andrea Griffiths and Daniel Adams from GitHub demonstrate how to harness GitHub Copilot directly on the GitHub website, revealing advanced AI-driven features and workflows for developers.

How to Chat with Your Repo & Create PRs with Copilot on GitHub

Presented by Andrea Griffiths and Daniel Adams

GitHub Copilot isn’t just an IDE companion; it powers enhanced AI-driven workflows right on the GitHub platform. In this video walkthrough, the presenters cover:

Key Features Demonstrated

GitHub Copilot Web Interface: How to use the GitHub Copilot dashboard on github.com/copilot.
AI Chat with Repositories: Ask contextual questions about your repository’s files, issues, and contents, and receive AI-generated answers within GitHub.
Issue Creation from Screenshots: Quickly generate new GitHub issues using screenshots and Copilot’s interpretation.
AI Control Center for Issues and Agents: Manage tasks and interact with AI-driven agents to streamline developer workflows.
Prototyping and Learning: Use AI to assist with prototyping ideas, exploring code, and enhancing your learning experience as a developer.
Switching Between AI Models: Select the best model for your task, depending on use-case and performance.
Thread Management: Organize and revisit AI-powered chat threads for specific project needs.
Differences: Copilot in IDE vs. On Web: Understand how Copilot’s capabilities are extended on the web, and when to use each mode.
Looking Ahead: AI-Native Development: Discussion of trends and the evolving role of AI in software development.

Takeaways

GitHub Copilot on the web extends developer productivity with interactive, repository-aware AI chat.
Capabilities include generating pull requests, responding to file-specific queries, and issue management from multiple input types.
Developers can easily switch between models and manage a history of AI conversations at the repository level.

Resources

Stay up to date by following GitHub on social media.