Pi is deliberately minimal.
Instead of forcing you into one fixed workflow, it gives you the freedom to choose your models, configure its behavior and extend it with exactly the features you need.
But that flexibility also means there’s a lot to explore: settings, security, context management, sessions, packages, skills, prompt templates, extensions and much more.
This brand-new course gives you a practical path from installing Pi and taking your first steps all the way to using it for real development work and building your own custom extensions.
In detail, you will learn:
How to install Pi and connect different AI models
How to configure Pi for global and project-specific usage
How Pi’s security model works - and how to run it in a Docker container for extra isolation
How to reference files, run terminal commands and manage context
How to use Pi for real project tasks
How to plan your work without a built-in Plan Mode
How to extend Pi with packages, guardrails and subagents
How to configure Pi with local and global AGENTS.md files
How to build and use agent skills and prompt templates
How to create custom Pi extensions
How to steer and queue work
How to clone, fork and navigate sessions
How to manage long-running sessions and context compaction
By the end of the course, you’ll be able to confidently use Pi for real development work, and shape it into the coding agent that fits
YOUR preferred workflow.
