Build and run agents you can see, understand and trust.
Your Personal AI Assistant; easy to install, deploy on your own machine or on the cloud; supports multiple chat apps with easily extensible capabilities.
Java framework for building distributed, long-running agents with an event stream, tool-call permissions, middleware, and sandboxed workspaces.
An open-source Collaborative Multi-Agent OS for transparent, human-in-the-loop task coordination via Matrix rooms.
ReMe: Memory Management Kit for Agents - Remember Me, Refine Me.
以AgentScope 为基础构建的企业级数字员工平台。
【融光】 - 基于 Agent 的全流程AI短剧/漫剧/视频创作平台 - Java & agentscope2.0 | Agent-based end-to-end AI creation platform for short dramas, motion comics, and videos – built on Java & & agentscope 2.0.
NovaNova Studio 是一个 Agent 驱动的 AI 创作工作台,集图片生成、视频生成、无限画布与多模型 API 、漫剧短剧生成于一体,为独立创作者和视觉团队打造更智能、更高效的创作体验。
OpenJudge: A Unified Framework for Holistic Evaluation and Quality Rewards
🤖 MateClaw — Your second brain with Multi-Agent Orchestration, MCP Protocol, Skills & Memory, Dream, and Multi-Channel Support. Built on Spring AI Alibaba.
Evidence-driven autonomous SOC team on AgentTeams: hash-bound approvals, HMAC audit chain, rollback and reproducible attack scenarios — auditable governance primitives for agent infrastructure.
【群策】 - 开源企业级云端多智能体协作 OPC 平台,组建你的专属 AI 团队 - Java & AgentScope 2.0 | [AgenTeam] Open-source, enterprise-grade OPC (One Person Company) platform for building collaborative AI teams – built on Java & AgentScope 2.0.
Trinity-RFT is a general-purpose, flexible and scalable framework designed for reinforcement fine-tuning (RFT) of large language models (LLM).
Build distributed, production-grade, long-running agents.
Typed RPC over a warm pool of PEP 734 subinterpreters — real parallelism, real error propagation, no hanging calls.
JSON.parse with hard limits on nesting depth, key count, string length and input size. Refuses parser bombs before they are allocated.
Express middleware that makes handlers safe to retry — replays the cached response for a duplicate Idempotency-Key instead of re-running it, with pluggable in-memory or Redis storage.
python sdk for dashscope [model studio](https://www.alibabacloud.com/en/product/modelstudio?_p_lc=1)
Auditable LLM multi-agent town simulation with AgentScope, Mesa, replay, policy experiments, and Streamlit
Cross-border commerce Agent prototype with retrieval, order tools, persistent memory, and observable execution