[!IMPORTANT]
Archive Notice
AgentScope 2.0 now ships a built-in Web UI as part of its batteries-included Agent Service — a FastAPI backend paired with a pre-built frontend (
examples/web_ui) for building, running and observing agents out of the box. Powered by a unified event stream, it visualizes agent reasoning, tool calls and multimodal content in real time, and offers multi-session isolation, Agent Team (leader–worker) orchestration with task planning, permission & human-in-the-loop control, IM channels (Feishu / Discord), MCP & Skill Hub management, session persistence and task scheduling.With these capabilities built natively into AgentScope 2.0, AgentScope Studio is no longer maintained as a standalone tool. We recommend all users migrate to AgentScope 2.0 (https://github.com/agentscope-ai/agentscope) for continued updates, new features and community support. This repository will remain available in read-only mode for reference and will be archived soon.
Thank you to everyone who contributed to and used AgentScope Studio!
AgentScope Studio: Development-Oriented Visualization Toolkit
[](https://arxiv.org/abs/2402.14034)
[](https://www.npmjs.com/package/@agentscope/studio)
[](./LICENSE)
[](https://agentscope.io/)
[](https://agentscope.io/)
[](https://agentscope.io/)
AgentScope Studio is a powerful local visualization toolkit designed for developers working with AgentScope. It provides a transparent, simple, and fun way to develop, debug, and evaluate agent-based applications.
Key Features
-
Project Management: Organize and manage your AgentScope projects with Projects and Runs
-
Runtime Visualization: Chatbot-style interface for real-time agent interaction
-
Tracing: OpenTelemetry-based trace visualization for LLM calls, token usage, and agent invocations
-
Agent Evaluation: Evaluation-oriented analysis from a statistical perspective
-
Built-in Copilot (Friday): A development assistant, playground for rapid secondary development, and integration hub for advanced features
📢 News
- [2025-08] AgentScope Studio is now open-sourced! We will continue to improve it and welcome contributions from the community.
💻 Installation
Prerequisites
- Node.js >= 20.0.0
- npm >= 10.0.0
- Docker (optional, for Docker deployment)
💡 Tip: If you're using nvm, you can run
nvm useto automatically switch to the required Node.js version.
# How to check your version
node --version # Should show v20.x.x or higher
npm --version # Should show 10.x.x or higher
From NPM (Recommended)
npm install -g @agentscope/studio
# Start AgentScope Studio
as_studio
From Source
git clone https://github.com/agentscope-ai/agentscope-studio
cd agentscope-studio
npm install
# Start in development mode
npm run dev
- From npm
npm install -g @agentscope/studio # or npm install @agentscope/studio
as_studio
Docker Deployment
Please refer to docker/README.md.
🚀 QuickStart
To connect AgentScope applications, you need to set the studio_url field in the AgentScope initializer as follows:
import agentscope
agentscope.init(
# ...
studio_url="http://localhost:3000"
)
# ...
📚 Documentation
For more details, please refer to our documentation:
- Overview - What is AgentScope Studio and how it works
- Quick Start - Installation and configuration guide
- Project Management - Managing projects and runs
- Tracing - OpenTelemetry integration and semantic conventions
- Friday - Built-in Copilot guide
- Contributing - How to contribute
⚖️ License
AgentScope Studio is released under Apache License 2.0.
✨ Contributors
All thanks to our contributors: