Rust framework for building modular, scalable LLM-powered applications. Joshua led the project at Playgrounds, setting technical direction, maintaining the framework, reviewing PRs, mentoring contributors, and helping move Rig from an early internal framework into a widely adopted open-source AI framework.

Current public proof: 7,542 GitHub stars, 839 forks, and 1,083,066 downloads for rig-core on crates.io as of Jun 6, 2026. Joshua reports it was around 450k downloads on his final day leading it.

Open source, actively maintained

An ergonomic, opinionated memory interface for AI agents, exploring how agent memory can be made easier to work with from Rust.

Public GitHub repository by joshua-mo-143 with 39 stars as of Jun 6, 2026.

Open source experiment

A minimum viable reproduction of Rust's tracing library, built to show how tracing works from the inside. The repo includes both a single-crate implementation and a split workspace architecture to make the design easier to study.

Public GitHub repository by joshua-mo-143, created May 2026, written in Rust, with README describing the tracing architecture.

Educational implementation

Example project showing how to create a Recursive Language Model with Rust and Rig.

Public GitHub repository by joshua-mo-143 with 67 stars and 7 forks as of Jun 6, 2026.

Example project

Python demo for a research agent on Venice AI, including an equivalent Deep Research report from Perplexity to benchmark effectiveness.

Public GitHub repository by joshua-mo-143, written in Python, with README/repo description documenting the benchmark comparison.

Demo project