Last Updated: 2026-03-26
MicroAgent is a local-first iPhone and iPad learning app for understanding how AI agents are built. It combines replayable demos, visual explanations, and source-oriented lessons so users can study agent behavior step by step instead of treating it like a black box.
What the app covers
- A visible agent loop from prompt to tool result
- Tool dispatch, planning, memory, and multi-step execution
- Collaboration patterns such as teammates, approval flows, and background work
- Task isolation, worktree-style thinking, and a polished end-state runtime demo
Learning format
MicroAgent is designed as a structured lesson experience rather than a generic AI chat app.
- Lessons use replayable in-app simulations so key behaviors stay stable and easy to inspect
- Source examples stay close to real runtime structure, making it easier to compare UI explanations with code
- The app includes a free starting section and a one-time unlock for the remaining structured lessons
- Open-source acknowledgements and license visibility are available from the app settings
Audience
MicroAgent is best suited to teens, students, developers, and curious builders who want to understand AI agent architecture in a concrete way. Early lessons stay visual and approachable; later lessons move into more advanced orchestration and coordination patterns.