open-source

An open-source reinforcement learning training loop for LLMs, spanning small CPU-friendly experiments like CartPole to async RLVR runs on GPU clusters. The repo highlights a compact codebase, quickstart commands, task customization, and support for algorithms like REINFORCE, PPO, GRPO, and RLOO.

An Electron desktop app for coordinating AI coding work by feature, using local Markdown cards and Git worktrees. It connects with Codex CLI and Claude Code CLI, tracks agent activity, commits, token usage, and keeps project context inside the repository.

Open-source coding agent for VS Code that can inspect a workspace, plan changes, edit files, run commands, and work with free providers or local Ollama models.

Open-source Go control-plane proxy for AI agents that enforces identity, policy, budgets, and audit for MCP and other upstreams, with anomaly detection that can auto-block compromised agents.

A Rust CLI that renders terminal text effects as a single static binary, ported with byte-exact parity from TerminalTextEffects. The README explains usage, build steps, supported platforms, and the project’s fidelity/performance goals.

Rust-native desktop cockpit for coordinating multiple AI coding tools like Claude Code, Codex, and Gemini CLI in one interface. The page shows installation steps, supported platforms, core workflows, safety defaults, and release notes.

Open-source governance harness for AI agents with policy enforcement, human approvals, isolated execution, budgets, and signed audit evidence. It includes a CLI, REST API, web dashboard, SDKs, and adapters for common agent frameworks.

Lucidos is a local AI OS and companion that keeps described apps running, responds to events, and works from a persistent desktop or browser-based interface. The repo shows install paths for macOS and Linux, a headless runtime, source build instructions, and a clear focus on user-owned local data.

Fast Rust MCP aggregator that exposes a small fixed tool surface for agents while semantically searching and calling tools across many downstream MCP servers. Supports SQLite-backed persistence, restart-safe embeddings, stdio and Streamable HTTP transports, plus a REST API for testing.

An AI-powered feedback widget and workflow for product teams. It captures screenshots, DOM snapshots, and console context, then routes feedback into an agent-friendly fix process with CLI, MCP, and skill support.

Open-source job search agent that searches multiple job boards, scores listings against a user profile with a local LLM, and includes a terminal UI plus a web dashboard.

Open-source shared memory layer for multi-agent systems, built around signed, cite-able facts that can be read without an account and integrated via MCP, Python, TypeScript, or REST.

Chrome extension that saves AI chat answers locally, adds tags, and lets users search and reopen them later.

Turo is a CLI proxy that reduces prompt text to content words to save tokens for coding agents. The repo shows install options, command examples, and support for multiple agents and modes.

CLI tool that hibernates and restores Claude Code sessions across reboots, with support for multiple terminal backends and exact session resume.

Open-source Python library for AI agent context governance with tamper-evident audit trails, provenance validation, freshness checks, and adapter support for common agent stacks.

Open-source control plane for managing Codex agents across machines, with a browser dashboard, event-driven task routing, and local setup instructions.

A terminal coding agent for Mac that runs on local models through MLX, edits code, runs tests, and keeps everything on-device.

Open-source AI app builder that turns a prompt into a full-stack Next.js app with hosting, sandboxing, database, auth, GitHub sync, multitenancy, and custom domains. The repo is positioned as a self-hosted alternative to v0, Lovable, Bolt, and Replit, and is meant to be embedded into another SaaS or run standalone.

Harness is an agentic IDE for running coding agents on remote machines over SSH, with one-click backend setup, backend switching, web access, and notifications across machines.

An open-source control plane for AI agents that adds policy enforcement, approval gates, durable execution, audit logs, and observability for unattended workflows.

A local-first harness for turning coding models into reliable software agents, with memory, sandboxed execution, verification, recovery, GitOps, and deployment workflows. The repo describes a multi-agent architecture, approval gates, provider support, and a real shipped example.

An open-source project that runs full games of Diplomacy between 7 LLM-controlled powers, with negotiation rounds, validated orders, and saved transcripts for replay and analysis.

Local-first federated search for AI agents that queries transcripts, files, a knowledge graph, a vector store, and the web, then merges results with trust-weighted ranking.