Skip to content

Memfun

Memfun is an open-source autonomous coding agent that writes, verifies, and reviews its own code.

Code Video Cerebral Valley

Core trick

Escalates coding tasks from a 2-LLM-call fast path to parallel multi-agent workflows, and uses DSPy recursive language models to read codebases beyond the context window.

Tags

dev toolsagents

Description

Memfun is an open-source autonomous coding agent that writes, verifies, and reviews its own code. It combines three ideas: a multi-strategy coding engine that escalates from fast 2-LLM-call solutions to parallel multi-agent workflows, DSPy Recursive Language Models for handling codebases beyond any context window, and a pluggable event-driven runtime that scales from a laptop to a distributed cluster. It features 9 specialist agents, a real-time web dashboard, persistent memory across sessions, and portable Agent Skills compatible with 20+ AI tools. Install in one command, run memfun init, and start building.
indoor47/memfun★ 93 forksPythonApache-2.0Created Feb 12, 2026 · last push Jun 25, 2026

README excerpt

<p align="center">
  <h1 align="center">Memfun</h1>
  <p align="center">
    <strong>An open-source autonomous coding agent built on three ideas:<br/>a coding engine that verifies its own work, infinite memory via Recursive Language Models,<br/>and scalable multi-agent orchestration on an event-driven runtime.</strong>
  </p>
  <p align="center">
    <a href="LICENSE"><img src="https://img.shields.io/badge/License-Apache_2.0-blue.svg" alt="License" /></a>
    <a href="https://python.org"><img src="https://img.shields.io/badge/python-3.12%2B-blue.svg" alt="Python" /></a>
    <a href="https://github.com/indoor47/memfun/actions/workflows/ci.yml"><img src="https://github.com/indoor47/memfun/actions/workflows/ci.yml/badge.svg" alt="CI" /></a>
    <img src="https://img.shields.io/badge/tests-854_pass-brightgreen.svg" alt="Tests" />
    <img src="https://img.shields.io/badge/packages-7-orange.svg" alt="Packages" />
  </p>
</p>

## Three Pillars

Memfun is built on three foundational ideas that work together. Each one solves a
hard problem in autonomous coding agents. Together, they produce something greater
than the sum of their parts.

<table>
<tr>
<td width="33%" valign="top">

### 1. The Coding Agent

A multi-strategy solver that writes, verifies, and reviews its own code. Starts
fast (2 LLM calls), escalates to parallel specialists when needed, and always
runs your linters before delivering results.

</td>
<td width="33%" valign="top">

### 2. Recursive Language Models

DSPy RLM

Team