Today we venture in the world of AI with Jason Pearson and the Android MCP (Model Context Protocol) SDK, a new library that allows Android developers to build apps that can communicate with AI models more effectively.
While this library is still in its early days, this solution has a lot of potential to revolutionize how we interact and debug our Android apps. Jason is going to share insights from his with MCP and what it enables for Android developers: from layout inspections, to network debugging and potentially automating and simplifying UI testing for mobile apps!
Enjoy the show 👨‍🍳
Show Notes
- 00.00 Intro
- 00.46 Episode Start
- 01.16 Jason’s Introduction
- 02.02 What is the Android MCP SDK?
- 04.36 MCP Use cases
- 08.19 Shipping MCPs in production apps
- 12.32 Installing the MCP SDK
- 14.36 Project Maturity
- 16.21 Building a new mental model
- 22.45 What got into building an MCP server?
- 25.24 How does the MCP protocol look like?
- 29.36 Which APIs from the MCP spec have you implemented?
- 31.18 What’s next for the Android MCP SDK?
- 34.49 The future of AI
- 40.42 Futher reading
- 43.55 Where people can find you online?
Resources
- kaeawc/android-mcp-sdk on GitHub
- Official Project Documentation
- Mentioned Resources:
- Why MCP really is a big deal - Model Context Protocol with Tim Berglund
- AI Code Assistants for Android Engineers
- @kaeawc on GitHub
- @kaeawc on Twitter
- @kaeawc.bsky.social on Bluesky
- Jason’s Personal Website