Stay up to date with the latest iOS CI/CD news

Every two weeks, directly to your inbox and packed with everything you need to know and be aware of so that you're always prepared for the next app release!

Read and trusted by engineers from:

Issue 66

Sponsored

Level Up Your iOS Code Reviews: Best Practices for High-Quality Pull Requests

Make your iOS development faster and smoother with smarter code reviews. Appcircle gives your team the tools to avoid bottlenecks and keep pull requests clean, clear, and quick to merge.

Enter the Automation MCP Server era!

Hey everyone! I hope you’ve all had a fantastic couple of weeks. Welcome to another issue of the newsletter!

In November of last year, Anthropic open-sourced the Model Context Protocol, a new standard for providing context to LLMs. Most of you will have heard of or even used MCP servers by now, but if you haven’t, let me explain why they are needed.

AI Assistants such as Claude or ChatGPT cannot access live data by default and are constrained to a set amount of information. There are good reasons why that’s the case, but as you can imagine, the power that access to live data or even the power to act on your behalf on certain services is incredible.

That’s where the Model Context Protocol comes into play. It provides an open-source, standard interface for connecting LLMs to live data sources. With the Open-Source SDKs, developers can now build servers that expose a set of tools that an AI assistant can invoke.

The community’s response to this has been incredible, and there are already numerous ready-to-use MCP Servers. For example, you could install the Linear, Filesystem, GitHub, and Slack MCP servers and feed your AI assistant the following prompt:

Find the first issue with an ‘In Progress’ status assigned to ‘Pol’, make the necessary changes to the codebase, create a Pull Request on GitHub, and finally, send the link to the new Pull Request in the #pull-requests channel on Slack.

This is where the true power of MCP Servers comes into play, as the AI assistant will be able to pick and choose from all installed tools by the conversation’s context and then parse the contents of the response to provide the best possible answer to the user 🚀.

In this issue, I will share some examples of MCP servers in real life and how they can be applied to CI/CD. If you have found a cool use case you’d like to share, feel free to reply to this email or reach out to me on social media!

💡 If you are interested in building an MCP Server but don’t want to write Kotlin, Python, or TypeScript (I don’t blame you 😅), check out this tutorial on how to build an MCP Server in Swift by Artem Novichkov.

🤖 Interacting with GitHub Actions from an AI Assistant

While scrolling through the list of MCP servers in the official Model Context Protocol repository, I came across this very interesting one: github-actions-mcp-server.

As its name states, the server exposes a set of tools that bring context as well as the possibility of performing tasks on your behalf in GitHub Actions.

🤖 Interacting with Bitrise from an AI Assistant

Bitrise have been super quick to build their own MCP server that exposes tools to get live data from their API. Similarly to the unofficial GitHub Actions MCP I shared above, the MCP server also brings functionality such as triggering or canceling builds to your AI Assistant of choice.

Read this amazing blog post from Arpad Kun to find out more about what you can do.

🍎 Can your AI Assistant help you with App Store Connect issues?

Last MCP server I share in this issue, I promise! But I thought this was cool and worth sharing! 😅

Joshua Riley built an MCP Server that exposes information from our beloved App Store Connect to your LLM of choice so that you can better diagnose issues and access live data for your apps and users 👏.

🚀 Is fastlane the right tool for Mobile Automation in 2025?

Fastlane is an open-source source established solution that many mobile apps use to automate their processes. While fastlane is still incredibly powerful and relevant, as it provides a vast amount of built-in actions, it is written in Ruby and requires developers to context-switch when working on their automation code.

What alternatives are there in 2025? Is there a good Swift-based solution? Pedro Piñera walks you through some of the options you can pick from.

📦 Inspect and secure your app’s .ipa files

Keeping an eye on the size and contents of your .ipa files is incredibly important to make sure that new releases don’t introduce unexpected changes to your final binary.

Artem Novichkov walks you through how to inspect .ipa files and explore their internal structure. Once you have read this and know what to look out for, make sure you automate the process on CI/CD! 🚀