agentscope-ai/agentscope

Project overview

AgentScope is a multi-agent LLM framework designed for production scenarios requiring fine-grained tool control, multi-tenancy, and isolated code execution. It provides developers with event-driven human-in-the-loop support, multi-tenancy serving, and extensible sandbox workspaces.

Project type

AI Agent · Infrastructure

Use cases

Automation

Deployment

Refer to project documentation

License

Apache-2.0

Best for

• Developers and AI engineers building multi-agent LLM applications.
• Enterprise teams requiring multi-tenancy and safe code execution environments.

Key capabilities

• Implements fine-grained, configurable control over tools and resources.
• Executes tools and code in isolated environments, featuring built-in backends for local, Docker, E2B, OpenSandbox, and Daytona.

Limitations and risks

• The framework requires an external LLM provider, such as the DashScope Model API used in examples, to operate.
• GPU requirements, minimum hardware specifications, data boundary configurations, and telemetry tracking are not documented.

Getting started

• Getting started involves a medium difficulty setup that requires coding knowledge. Users interact with the framework primarily as a library.

Evidence and sources

• GitHub project description: Build and run agents you can see, understand and trust.
• README: AgentScope 2.0 is a production-ready, easy-to-use agent framework with essential abstractions that work with rising model capability and built-in support for .
• Release: v2.0.4.post1
• README: [2026-07] `INTE`: Daytona-based workspace/sandbox supported.
• README: An extensible FastAPI based **multi-tenancy**, **multi-session** agent service with pre-built Web UI in `examples/web_ui`

Reviewed at: 2026-07-17T10:09:10Z