agentscope-ai

TypeScript implementation of AgentScope, a framework purpose-built for multi-agent systems.

A platform for running AI agents on real-time sports data and making predictions about game outcomes.

Multi-tenant fine-tuning framework for local LLMs with a Tinker-compatible API. Supports SFT and RL training.

Curated collection of skills around the AgentScope ecosystem and CoPaw applications, compatible with Claude Code and Cursor.

Java runtime for deploying agents and running them inside a tool sandbox — the JVM port of the AgentScope framework.

UI component library for building AgentScope and Alibaba Cloud Apsara Lab agent interfaces.

Development-oriented visualization toolkit for AgentScope. Trace, debug, and monitor multi-agent LLM systems with an OpenTelemetry-aware UI.

Unified framework for holistic LLM evaluation and quality rewards, with reward models and grader skills for RLHF and agent alignment.

General-purpose, scalable framework for reinforcement fine-tuning of large language models, with RLHF support.

Production runtime for agent apps with secure tool sandboxing, Agent-as-a-Service APIs, and full-stack observability.

AgentScope's memory-management kit for AI agents — episodic, semantic, and procedural memory stores with retrieval and consolidation.

AgentScope's agent-oriented programming framework for the JVM — define agents, message passing, and orchestration in idiomatic Java.

Self-hosted personal AI assistant from AgentScope, easy to deploy locally or in the cloud, with chat-app integrations and extensible capabilities.

A production-ready framework for building, deploying, and managing AI agents with built-in ReAct agents, multi-agent orchestration, voice support, and finetuning capabilities. Designed to work with modern LLMs while providing essential abstractions for agent development.