Copilot Code Review Gains AGENTS.md Support and Enhanced UI

GitHub has rolled out key enhancements to its Copilot code review functionality, focusing on deeper customization and improved user experience. A major highlight is the introduction of support for `AGENTS.md` files at the root of a repository. This new capability empowers development teams to provide Copilot with specific, repository-level instructions, ensuring that the AI-generated code reviews are more relevant and consistent with the project's established guidelines and coding standards. When an `AGENTS.md` file is present, Copilot automatically incorporates its context into the review process, leading to more precise and actionable feedback. Beyond the backend intelligence, the user interface for Copilot code review has also seen notable improvements. Developers can now more easily initiate a Copilot review on draft pull requests through a dedicated 'Request' button, which appears directly within the reviewer picker. This streamlines the workflow, eliminating the need to search for Copilot manually and making it quicker to get initial AI feedback before a pull request is finalized. Furthermore, GitHub has addressed potential visual clutter within pull request timelines. Previously, Copilot's review events could sometimes overwhelm the conversation tab. To mitigate this, certain Copilot code review timeline events are now intelligently collapsed. This change helps declutter the interface, allowing developers to focus on critical discussions and changes without being distracted by an excessive number of AI-generated entries. These UI adjustments collectively aim to enhance the overall efficiency and clarity of the code review process when leveraging Copilot.