Havuz

Tinybot is a lightweight personal AI Agent that is constantly evolving

Android Harness AI Coding Agent Native app for mobile, No Root No PC

智能体 = 大模型 + Harness。深入剖析 Harness 工程原理、设计、实现与实践！

零远程行为、子代理协同、支持自主进化的自治 Agent 编队丨 Autonomous Agent Fleet with Zero-Remote Behavior: Sub-agent Collaboration and Self-Evolution

AI coding agent for Visual Studio 2022/2026 — chat, inline completion, and five autonomous agents that explore, plan, patch, and verify buil

An open-source agent harness for software work.

Graph-Orchestrated Agent Loop — a production-grade framework on LangGraph. Combine workflow graphs and agent loops, transpile Dify DSL to ru

jev-use, macOS’ta sesli veya yazılı komutları Accessibility tree üzerinden uygulama eylemlerine dönüştüren bir computer-use aracı.

Neden yükseliyor · README, vision model kullanmadan Accessibility tree ile bilgisayar kontrolü vaadini ve hızlı kurulum akışını öne çıkarıyor. Ölçümlerde repo 8 günlük, 100 yıldızlı; son 28 günde 8 commit ve 2 farklı yazar var. Bu, erken ilgi ve bir miktar geliştirme faaliyeti gösteriyor ancak benimseme veya olgunluk kanıtı değil.

HAR: open agent harness (CLI + MCP) for coding agents. Isolated worktrees, deterministic verify, software factory workflows for Claude Code,

A benchmark for evaluating LLM × harness performance.

Anaxa 是一个面向科研工作流的开源智能体系统。它不是单纯的聊天机器人，也不是无人监管的自动发论文机器，而是把文献检索、证据审计、实验执行、论文写作、同行评审式检查和最终产物打包放进同一个可追踪的研究生命周期中。

Open-source enterprise AI workforce platform — containerized roles, declarative skills, MCP tools, policy-driven security, K8s-native schedu

Official Code for the paper "Doc2Harness: Benchmarking LLMs on Long-Horizon Harness Engineering of Coding Agents "

QuantClaw is a plug-and-play task-type routing quantization plugin for OpenClaw.

OpenCode++: a Coding Agent Reliability Harness for OpenCode, adding context, edit boundaries, command evidence, verification gates, impact a

AI framework for Ruby/Rails developers

Aetox harness: a local-first Windows AI environment - model, skills, team and memory as one system. Two doors - an Assistant that adapts to

Local multi-agent harness for running persistent agent teams, long-run, society-simulation

AI智能体系统

Enables LLM to use a cli environment.

self-extending sandboxed agentic harness. a loop, two tools and an llm

Remote coding harness exposed as a secure Streamable HTTP MCP server

Agent Harness Platform — shareable workspaces, MCP tools, skills, memory, and cron/webhook automations. Self-hosted and transparent.

基于 Deepseek-Harness 核心底座打造的桌面版 Agent.继承底坐全部能力。并补全 IDE 相关功能。

Cross-platform AI agent harness for 12+ hour autonomous development cycles. Spec-driven, DAG-orchestrated, with a 3-critic evaluation pipeli

An Orchestrate multiple coding agent desktop app

A minimal local coding agent in Go — a faithful port of Sebastian Raschka's mini-coding-agent, with no LLM framework.

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

Fa - Factory Agent. Cross-platform AI agent harness for Dart and Flutter — streaming providers, agent loop with tools, session persistence,

Open-source desktop AI coding platform with built-in Pi Agent (pi-coding-agent) — a graphical AI programming assistant that lets non-technic

AgentScope Java, Java ile uzun süre çalışan ve dağıtık ortamlara taşınabilen agent uygulamaları geliştirmek için bir framework sunuyor.

Neden yükseliyor · 13 günlük yaşına rağmen 108 yıldıza ve son 28 günde 12 commite ulaşmış; Java tarafında üretim odaklı agent altyapısı arayanların ilgisini çekmiş olabilir. Ancak yıldız ve README özellikleri benimseme ya da üretim başarısının kanıtı değil; farklı yazar sayısının 1 olması katkı tabanının şimdilik dar olduğunu gösteriyor.

Lightweight Claude Code alternative: ~1.8% lines, ~90% capability

An open source, self-hosted, privacy-focused AI workbench

A human-governed AI coding workflow that distills ephemeral session context into persistent project memory—making work traceable, reviewable

Continual harness optimization

One-shot LLM eval cases by NAGI STUDIO - same prompt, different agents (model + harness), runnable artifacts side by side.

Map, inspect, and audit AI-agent markdown ecosystems (skills, agents, commands, hooks) as a graph.

Runs, traces, tool calls, LLM calls, errors, latency, tokens, and eventually detection for agents that appear stuck or caught in loops.

Spec-driven development and context engineering for Claude Code, Cursor, Codex, and GitHub Copilot — backed by project context in Git.

Craft is a Claude Code plugin that asks you the product questions and works out the technical ones. Features get planned, built, and tested,

Long-term, cross-project memory for AI coding agents. Your own Obsidian vault as the source of truth. Daemonless and without opaque database

Most AI agents forget you the moment the tab closes. Constellation Engine gives them a hippocampus — a living star map with spreading activa

A token efficient, local LLM friendly agentic coding harness: a complete agent runtime (agent loop, tool pipeline, sandboxing, session persi

A harness for embodied intelligence — connecting agents, robot skills and the physical world through execution, feedback and recovery.

Tools for testing APIs with AI agents

Elixir-native LangChain, LangGraph, and DeepAgents for traceable LLM apps: OTP workflows, tools, memory, human-in-the-loop, streaming, custo

Open-source Environment toolkit of claw-like agents, support task/harness generation and evaluation

Hive Agents harnest agent loop localfirts

👋 Claude Code & Codex harness insights dashboard — visual catalog and token analytics for skills, agents, plugins, hooks, memory, and more!