mini-pi is a minimal AI coding agent implemented from scratch in Python. It is designed primarily as a learning tool to help developers understand the inner workings of autonomous coding agents that use LLMs. The CLI-based package uses models such as DeepSeek and is fully open source.
mini-pi sits in PulseGate's AI & ML category. It focuses on understanding how AI coding agents work by studying and experimenting with a minimal implementation. It is built as an open-source project for developers. The project is open source (MIT). It ships for the command line.
uonxhou builds and maintains mini-pi, and it first shipped in 2026. Development happens publicly on GitHub with 7 commits in the last 90 days. Among its 3 catalogued features are AI coding, autonomous agent, and learning tool.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do