The post GitHub Copilot for Xcode: Unlocking Agentic Power for Apple Developers appeared first on Xcode.
]]>2025 marks the rise of intelligent agents. Developers need flexibility and precision—both in how they interact with agents and how those agents adapt to their coding style. After launching Custom Agents in VS Code, we’re now bringing them to Xcode with Custom Agents, Subagents, and the powerful Plan Mode.
Configure Copilot to match your workflow. Define instructions, constraints, and tools so Copilot works the way you do—whether you’re building SwiftUI interfaces or managing complex iOS projects.
Each subagent runs in a clean, focused context, reducing noise and improving accuracy. This reduces distractions and improves reasoning for actions like fixing tests, refactoring code, or generating documentation. In Agent Mode, Copilot automatically invokes your pre‑defined subagents when appropriate, so the right capability is applied at the right time—no manual routing required.
Plan Mode transforms Copilot into a strategic problem-solver. It breaks down multi-step challenges into clear, sequential actions—planning, executing, and validating—so you can tackle big engineering tasks with confidence.
Stay in flow with proactive recommendations for your next code edits, improvements, or cleanup—helping you iterate faster without losing context.
In addition, GPT‑5.1 models, including GPT‑5.1‑Codex and Codex‑Mini, are now available in Copilot for Xcode. These models bring significant enhancements to chat responsiveness, agent-driven workflows, and code generation accuracy—helping developers plan, execute, and iterate more efficiently within Xcode.
Try it today in the stable version of Copilot for Xcode. You will also need to have a valid Copilot license. Get started with our official documentation.
We’d love to hear your thoughts! Share feedback via the GitHub Copilot for Xcode Feedback channel to help us continue improving the product.
The post GitHub Copilot for Xcode: Unlocking Agentic Power for Apple Developers appeared first on Xcode.
]]>The post MCP Registry and Allowlist Controls Now Available in GitHub Copilot for Xcode (Public Preview) appeared first on Xcode.
]]>An MCP Registry is a directory of Model Context Protocol (MCP) servers. With Copilot in Xcode, you can now browse, install, or uninstall MCP servers from the registry —making setup fast and seamless.
Admins can configure which MCP servers are available to developers and enforce registry-only mode when needed.
Note: This is an early public preview and may evolve based on feedback. Current enforcement uses server name matching; additional validation and more granular policies are planned. For details, see the official docs: Configure MCP server access.
In Xcode:
Allowlist Controls are available for Copilot Business and Copilot Enterprise customers.
Try it today in the pre-release version of Copilot for Xcode. You will also need to have a valid Copilot license. Get started with our official documentation.
We’d love to hear your thoughts! Share feedback via the GitHub Copilot for Xcode Feedback channel to help us continue improving the product.
The post MCP Registry and Allowlist Controls Now Available in GitHub Copilot for Xcode (Public Preview) appeared first on Xcode.
]]>The post Model Context Protocol (MCP) for GitHub Copilot in Xcode is Now Generally Available appeared first on Xcode.
]]>With MCP support in Xcode, developers can now connect Copilot to both local and remote MCP servers that follow the MCP specification in your IDE. Once configured, Agent Mode can leverage tools exposed by these servers to assist with broader tasks. For example, you can configure a remote GitHub MCP server using a PAT or OAuth to authenticate. Enable actions like creating issues, checking repo history, and searching GitHub. Here is a short demo for it.
For organization or enterprise administrators, MCP access is controlled by the MCP servers in Copilot policy. The MCP policy is disabled by default and must be enabled by an organization or enterprise administrator.
To use MCP, install the latest GitHub Copilot plugin for Xcode. You will also need to have a valid Copilot license. Get started with our official documentation.
We’d love to hear your thoughts! Share feedback via the GitHub Copilot for Xcode Feedback channel to help us continue improving the product.
The post Model Context Protocol (MCP) for GitHub Copilot in Xcode is Now Generally Available appeared first on Xcode.
]]>The post GitHub Copilot for Xcode: Introducing Copilot Vision, Custom Instructions, Locale-Aware Responses, and More appeared first on Xcode.
]]>You can now upload images—such as screenshots, diagrams, or UI mockups—directly into Copilot Chat. Whether it’s a screenshot of an error, a diagram, or a UI mockup, Copilot Vision helps you understand and reason about visual content right inside Xcode. To help you get started, here is a quick demo—building a classic Snake game in Swift, right inside Xcode.
Customize how Copilot responds to you! You can now configure custom instructions at both the global level and per workspace. Tailor Copilot’s behavior and tone to better suit your coding style and project needs.
Copilot Chat now supports locale-aware replies. Set your preferred language in the GitHub Copilot for Xcode settings, and Copilot will respond in the language that best fits your development context—making collaboration across global teams even smoother.
The Copilot Chat window now auto-attaches to your Xcode workspace and can be toggled in settings, providing a more integrated and efficient workflow.
To access GitHub Copilot for Xcode, you’ll need a Copilot license . Simply sign in with your GitHub account and start using AI-powered assistance in your Xcode IDE.
You can read more in our official documentation.
Your feedback is essential in shaping the future of GitHub Copilot for Xcode! Share your thoughts our feedback channel or join the GitHub Community to discuss your experiences and provide suggestions.
Try it out today, explore what’s possible, and let us know what you think. We can’t wait to see what you build.
The post GitHub Copilot for Xcode: Introducing Copilot Vision, Custom Instructions, Locale-Aware Responses, and More appeared first on Xcode.
]]>The post Copilot for Xcode Adds Support for Anthropic Claude Sonnet 4 and Opus 4 appeared first on Xcode.
]]>
For full details, please see this documentation.


Your feedback drives improvements. Let us know what you think using the in-product feedback option, or share your thoughts in this channel. We’d love to hear how Copilot is helping your Apple development workflow.
The post Copilot for Xcode Adds Support for Anthropic Claude Sonnet 4 and Opus 4 appeared first on Xcode.
]]>The post GitHub Copilot: Exploring Agent Mode and MCP Support in Public Preview for Xcode appeared first on Xcode.
]]>Agent Mode turns GitHub Copilot into an intelligent pair programmer that can understand your intent, plan multi-step tasks, and execute them—all through natural language. Unlike traditional autocomplete, Agent Mode is designed for more complex workflows. You can switch to Agent Mode when your task involves multiple steps, needs to integrate with external systems (like an MCP server), or when you’d rather let Copilot figure out the best way to get it done.
In Xcode, Agent Mode can:
Here’s a quick demo of Agent Mode in action. Let’s see how it helps add new features to the HomeView, and adapts to support a new UX toggle while developing an iPhone app. All of this is done automatically, from a single prompt.
Model Context Protocol (MCP) is an open standard that allows AI models to interact with external tools and services. In Xcode, MCP extends Agent Mode by enabling it to invoke tools contributed by MCP servers—unlocking workflows like querying file metadata, running diagnostics, or integrating with custom services.
Tool Picker in Agent Modemcp.json)You may use this JSON file as a starter:
{
"servers": {
"memory": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-memory"
]
}
}
}
Once configured, you can use natural language to call tools exposed by your MCP server. For example, you can ask the agent to get the last modified time of the current files, and it will route the request through the appropriate tool.
Agent Mode and MCP support are now available in public preview for all Xcode users. To get started, install or update to the latest version of GitHub Copilot for Xcode. To access GitHub Copilot for Xcode, you’ll need a Copilot license . Simply sign in with your GitHub account and start using AI-powered assistance in your Xcode IDE.
You can read more in our official documentation:
Your feedback is essential in shaping the future of GitHub Copilot for Xcode! Share your thoughts our feedback channel or join the GitHub Community to discuss your experiences and provide suggestions.
Try it out today, explore what’s possible, and let us know what you think. We can’t wait to see what you build.
The post GitHub Copilot: Exploring Agent Mode and MCP Support in Public Preview for Xcode appeared first on Xcode.
]]>The post GitHub Copilot for Xcode: Explore @workspace and New Models Support (GPT-4.1, GPT-4.5 and More) appeared first on Xcode.
]]>@workspace context: By referencing @workspace in Copilot Chat, you can ask questions about your entire codebase in Xcode. GitHub Copilot will analyze your project’s codebase and provide detailed answers, including references to relevant files and symbols. The @workspace context helps you find the code you’re looking for, understand how functionality is implemented, and much more. Here is a short demo of it.
New model support: New models including OpenAI GPT-4.1, GPT-4.5, o3, o4-mini, Gemini 2.5 Pro and Claude 3.7 Sonnet are now available in the model selector of Copilot Chat. These models bring enhanced capabilities and improved performance to your coding experience. Learn more about the models available in Copilot in our documentation.
This update also includes a few other UX improvements and bug fixes:
To access GitHub Copilot for Xcode, you’ll need a Copilot license . Simply sign in with your GitHub account and start using AI-powered assistance in your Xcode IDE.
Additionally, GitHub Copilot offers a free tier, which includes 2,000 code completions and 50 chat messages per month. If you don’t have a GitHub account yet, you can create one and start exploring Copilot’s capabilities today.
You can read more in our official documentation.
Your feedback is essential in shaping the future of GitHub Copilot for Xcode! Share your thoughts using the in-product feedback option or join the GitHub Community to discuss your experiences and provide suggestions.
Join us on this journey as we continue to enhance GitHub Copilot for Xcode and deliver a seamless AI-powered coding experience!
The post GitHub Copilot for Xcode: Explore @workspace and New Models Support (GPT-4.1, GPT-4.5 and More) appeared first on Xcode.
]]>The post GitHub Copilot for Xcode Chat – Public Preview Now Available appeared first on Xcode.
]]>GitHub Copilot is an AI-powered assistant that helps developers write code faster and with greater accuracy. Beyond code completion, GitHub Copilot for Xcode now offers intelligent suggestions for your specific coding tasks through interactive chat. Whether you need help debugging, understanding code, or exploring alternative approaches, Copilot’s chat provides real-time insights and guidance, streamlining your workflow and enhancing efficiency.
After installing GitHub Copilot for Xcode, you can open the chat in these two ways:
1. In the macOS menu bar > GitHub Copilot app menu > Open Chat
2. In the Xcode menu -> Editor -> GitHub Copilot -> Open Chat
Interact with Copilot directly in the chat view. Ask for help with coding tasks, request explanations, or get AI-driven code suggestions in real time. The chat interface provides an intuitive way to communicate with Copilot and receive relevant responses based on your queries.
Speed up your development process with slash commands. Use quick commands like /explain to get instant explanations for code snippets. These commands allow developers to efficiently gather insights without breaking their coding flow.
To help Copilot provide the most relevant answers, you can attach files to your chat prompt. Simply click the attach icon and select the files you want to reference.
You can maintain multiple conversation threads while working across different Xcode IDEs running simultaneously. Whether you’re debugging, building new features, or optimizing performance, Copilot helps keep discussions organized so you never lose context.
Easily track past interactions with Copilot. The chat history feature helps you revisit previous conversations, reference past suggestions, and streamline collaboration by keeping a record of your AI-assisted discussions.
To access GitHub Copilot for Xcode, you’ll need a Copilot license . Simply sign in with your GitHub account and start using AI-powered assistance in your Xcode IDE.
Additionally, GitHub Copilot offers a free tier, which includes 2,000 code completions and 50 chat messages per month. If you don’t have a GitHub account yet, you can create one and start exploring Copilot’s capabilities today.
You can read more in our official documentation.
Your feedback is essential in shaping the future of GitHub Copilot for Xcode! Share your thoughts using the in-product feedback option or join the GitHub Community to discuss your experiences and provide suggestions.
Join us on this journey as we continue to enhance GitHub Copilot for Xcode and deliver a seamless AI-powered coding experience!
The post GitHub Copilot for Xcode Chat – Public Preview Now Available appeared first on Xcode.
]]>