Standardize benchmark wrapping so the community can wrap various otherwise-incompatible benchmarks uniformly and use them everywhere.

Project Tapestry aims to give every nation and participant frontier AI they can call their own — uniting a global consortium to train a shared frontier model from which partners build and own sover…

Fast Tokens

ModSleuth builds evidence-grounded dependency graphs for LLM releases from public artifacts such as model cards, dataset cards, technical reports, release blogs, and code repositories.

Omnigent is an open-source AI agent framework and meta-harness: orchestrate Claude Code, Codex, Cursor, Pi, and custom agents — swap harnesses without rewriting, enforce policies and sandboxing, an…

Interactive subagents for pi — spawn, orchestrate, and manage sub-agent sessions in cmux terminals

CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies

Pi extension that optimizes RTK command rewriting and tool output compaction for the coding agent.

File Storage & Sandbox Backends for Pydantic AI: console tools for file operations, Docker-isolated sandboxes for safe execution, and permission system with presets for access control. Enables secu…

Lightweight and portable LLM sandbox runtime (code interpreter) Python library.

The AI Toolkit for TypeScript. From the creators of Next.js, the AI SDK is a free open-source library for building AI-powered applications and agents

A lightweight, powerful framework for multi-agent workflows

Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 60-95% fewer tokens, same answers. Library, proxy, MCP server.

Self-hosted API for your RL Infrastructure

Open-source, community-driven agent harness

Codified Context: Infrastructure for organizing project knowledge for AI coding agents

Data recipes and robust infrastructure for training AI agents

A zero-boilerplate games programming framework for Python 3, based on Pygame.

CheetahClaws: A Fast and Easy-to-Use Agent Harness Infrastructure for Long-Horizon, Multi-Model, and Tool-Using AI Systems

Code for the blog "Neural audio codecs: how to get audio into LLMs"

GPU accelerated population genetics statistics package

mini cli search engine for your docs, knowledge bases, meeting notes, whatever. Tracking current sota approaches while being all local

Local-first session search, analytics, insights, and token use statistics for coding agents, supporting Claude Code, Codex, and more than 20 other agents.

Experiments and examples using the Marin framework

npx ccusage

mKernel: fast multi-node, multi-GPU fused kernels

Agentic Schema for Transparent Research Analysis.

DeepSeek 4 Flash and PRO local inference engine for Metal, CUDA and ROCm

DeepSeek-native AI coding agent for your terminal. Engineered around prefix-cache stability — leave it running.