Havuz

:dog: Beagle is designed to be the unified platform for the broader Recursive Self-Improvement (RSI) ecosystem. Think of it as Hugging Face

Tuvis(兔维斯) — desktop mission control for AI coding agents: live session monitoring, remote injection, Skill/MCP management & usage ledger. O

Local-first AI companion daemon — an agent harness with its own long-term memory. Remembers you, grows with you, extends and repairs itself.

Claude Code 源码深度研究，包括 Foundations/Execution/Infrastructure 三大章节和 23 个子系统的架构分析拆解。

A living 3D city for your AI agents, self-hosted. Bring OpenClaw and Hermes agents, connect a Grok Bot, or build agents on any AI model: the

CubeLoop — a Pythonic, async-native agent harness core designed for high performance, readability, and production-grade persistence.

A pure-Go port of pi (earendil-works/pi) — the pi-ai unified multi-provider LLM API, pi-agent-core agent loop, and pi-coding-agent, native i

EOC: open-source operating system for OpenAI Codex workflows with agents, skills, hooks, rules, memory, safety gates, and cross-harness adap

Runtime-neutral AI execution bridge for Claude Agent SDK, Codex app-server, and Agent Client Protocol (ACP).

A local-first coding agent that doesn't take "done" on faith: the engine keeps the run going until the project's own tests pass, and every g

Build a coding agent from scratch, one harness primitive at a time. 15 chapters (ch-00..ch-14), each green on two gates.

A coding agent harness built specifically for Motif-3

The open-source agent harness, ready out of the box and customizable in depth. Tools, context, durable execution, permissions and recovery f

Benchmark for comparing agent harnesses on everyday online tasks — fixes the base model, varies the harness. Sister project of ClawBench, sa

Durable execution framework for AI agents in Go. Keeps agent state, tool calls, and execution loops resilient across process crashes.

VS Code extension kanban board for agentic AI workflows

An open-source, coding agent harness designed specifically for local LLMs. Designed as everything a user needs to jumpstart their own harnes

Assessment-driven, local-first learning agent built on an observable Agent Runtime and eval harness—grounded ingestion, trace replay, HITL a

An open-source AI agent for the browser. Bring any model. It plans in the side panel, acts on real pages through tools, and verifies every c

Verifiable AI coding agent harness / 可验证的 AI 编码代理：evidence hash-chain · hollow-answer gate · manual verification checklist · 6 engines + MCP

A terminal-native AI agent runtime built in Rust. Interactive chat, parallel agent orchestration, and autonomous supervision.

A lightweight, general-purpose harness for tool-using LLM agents, fair benchmark evaluation, harness baselines, and personal assistant workf

Pi adapter for official DeepSeek Harness minimal mode. V4 Pro thinks We need… / I need…, not Let me…

Unofficial, open-source Windows desktop harness for Kimi Code CLI.

Fable-style spec + evidence gate for Claude Code + Codex. Makes Opus/Codex work under Fable-like discipline: blocks every edit until a deter

Write, validate, and review Agent SKILL.md's directly in Visual Studio Code. SKILL.md Inspector provides live diagnostics for every file na

Skill repository to enforce Engineering disclipline in coding assistants

A minimal coding-agent harness built to measure what context management actually costs: constraint survival after compaction (8%→83%), def

Benchmark coding agents and agent harnesses on 226 deterministic file-editing tasks with byte-exact verification and a public leaderboard.

开源 · 本地优先的 AI 智能体：编程与实干兼备。 自带模型密钥，内置 MCP、技能、插件、工作流、知识库以及数据库。

Compact AI agent runtime control plane with governed tools, HITL approval, resumable execution, trace-driven evaluation, and SWE-bench-shape

A custom coding harness for TypeSafe AI's Jev: an LLM proposes, Jev answers narrow questions, code decides, every step leaves a receipt.

Fast, deterministic document parser and chunker for RAG and agent pipelines. Ships an MCP server for agents navigating long documents with p

Runtime for coding-agent harnesses. Run Claude Code, Codex, OpenCode, Pi, and Grok in one terminal; hop live between them; search or resume

Simple Long Horizon Agent - A simple yet effective AI agent for learning, experimentation, and long horizon work.

Safe to explore. Free to build. Koder by aixplain is an AI coding harness for desktop and terminal, with permission checks, reviewable chang

A minimal, evidence-gated harness for Codex and Claude Code: reuse what works, add only what is missing, verify before evolving.

Provider-neutral bounded agent runtime for Android/JVM: controlled context (trust/priority/budgets), profile-scoped tools, OpenAI-compatible

Nausicaa treats an Agent run as a dynamic topology of Lanes rather than one fixed linear loop. The model can decide when to observe, fan out

⚡ Millisecond System-1 judgement for every tool call in DeepSeek Harness — Jev-powered risk classification & evidence-gated auto-approval. F

Execution posture harness for Claude Code. Your coding agent stops one step early; Castra is the step.

An AI agent plays Baldur's Gate 3 on Linux through a Lua mod and an MCP server, seeing only what a player sees. Includes a model benchmark a

A minimal, language-agnostic harness for Claude Code and AGENTS.md-compatible coding agents. | Claude Code와 AGENTS.md 호환 에이전트를 위한 미니멀·언어

Agentic PR reviews for the IBM Carbon Design System, run by IBM Bob or Claude from a codified review prompt and grounded by Carbon MCP. Buil

AI-assisted XAUUSD daily research workflow with Research Skill, Agent Harness, LangGraph, Tool Calling, and free public data sources.

DSH 插件：把 WorkBuddy 桌面 App 里的模型接入 DeepSeek Harness，零配置直接用。（原生嵌入"插件-插件配置页"）

Make every model cheaper or better. Zero-config behavioral harness for Claude Code, with a reproducible 408-run benchmark

Open & nano reimplementation of Sakana Fugu. A repo-native, multi-agent coding loop powered by 9+ LLMs (isolated via Claude Code) and an ind

Declarative, local-first runtime for bounded AI agents — define an agent in one manifest, run it headless, and audit every effect.

Build the memory your agent actually needs — automatically.