AgentScope-AI
13 Tools
AgentScope is an open-source framework for building, evaluating and deploying AI agents, and agentscope-ai is the GitHub organization behind it and related projects such as AgentScope Studio and the QwenPaw personal assistant. The framework was introduced in a 2024 arXiv paper by Alibaba Group researchers. AgentScope 2.0 supports protocols such as MCP and A2A and can deploy agents locally, as serverless functions or on Kubernetes.
Tools
Agent framework with built-in ReAct agents, multi-agent orchestration, voice support, and model finetuning.
Self-hosted personal AI assistant from AgentScope with chat-app integrations, deployable locally or in the cloud.
Agent-oriented programming framework for the JVM covering agents, message passing, and orchestration in idiomatic Java.
Memory-management kit for AI agents with episodic, semantic, and procedural stores plus retrieval and consolidation.
Evaluation framework that scores LLM quality with reward models and grader skills for RLHF and agent alignment.
Traces, debugs, and monitors multi-agent LLM systems through an OpenTelemetry-aware visual interface.
Curated skills for the AgentScope ecosystem and CoPaw apps, compatible with Claude Code and Cursor.
General-purpose Python framework for reinforcement fine-tuning of large language models, with RLHF support.
Runs agent apps in production, sandboxing their tools and exposing them through Agent-as-a-Service APIs.
UI component library for building agent interfaces for AgentScope and Alibaba Cloud Apsara Lab.
Multi-tenant fine-tuning framework for local LLMs with a Tinker-compatible API and support for SFT and RL training.
TypeScript implementation of the AgentScope framework for building multi-agent systems.
JVM port of the AgentScope runtime for deploying agents and running their tools inside a sandbox.