> **A TypeScript Framework for Building Tool-Using AI Agents**

AgentX — Decentralized AI Agent SDK with ReAct AgentLoop + Multi-Tenant LLM Providers

Drop-in Anthropic SDK wrapper for AgentLoop. Adds memory retrieval and turn logging to every messages.create call.

Drop-in OpenAI SDK wrapper for AgentLoop. Adds memory retrieval and turn logging to every chat.completions.create call.

JavaScript / TypeScript SDK for AgentLoop — middleware that turns human corrections into searchable memory for your agent.

The concrete agent loop plugin for the DeepSeek Harness

A local implement-and-review loop that runs Claude and Codex against one task at a time, and publishes only what an independent audit has approved.

AI agent that monitors chat platforms and responds using Claude Code

Official SDK for the AgentLoop Agentic Marketing Network

AgentLoop integration for LangChain — earn up to 70% revenue share when your agent recommends relevant products.

OpenClaw memory plugin: long-term memory via Alibaba Cloud CMS AgentLoop Memory API .

AgentRuler SDK — AgentLoop core with Node.js backend adapter, permission system, and full toolbar settings API.

English | [简体中文](README-CN.md) 

Shared prerequisite mounting for tests that exercise the concrete agent loop

Opt-in durable per-step context with the current time and elapsed time

Agentic loop system with tool execution for LLMs

Adaptive multi-agent orchestration for AI agents: route work to typed specialist agents, learn from outcomes, mine patterns, and run propose-dispatch-verify-retry agent loops. Zero-dependency TypeScript library + CLI with a Claude Code hook generator.

Hyacinth 🪻 — multi-provider AI agent with tool system and context management

Leaving Blank Space Agent — 多 Agent 编排框架

YUAN Agent Core Runtime