agentUniverse
ActiveDescription
An LLM-based multi-agent framework that lets developers easily build multi-agent applications with core abstractions for agent roles, tools, knowledge management, and collaboration patterns.
Key Features
- PEER multi-agent collaboration pattern for Plan, Execute, Express, and Review workflow orchestration
- DOE pattern combining Data-fining, Opinion-inject, and Express agents for data-intensive tasks
- Wide LLM provider support including Qwen, DeepSeek, OpenAI, Claude, Gemini, Llama, and more
- Domain experience integration enabling smooth embedding of industry expertise into agent workflows
- Visual agentic workflow platform for low-code agent application development
- Standard project scaffolding with quick guides for building single and multi-agent applications
Use Cases
💡 Building domain-expert-level intelligent agents for financial industry analysis
💡 Creating multi-agent systems for event interpretation and report generation
💡 Developing automated financial report pipelines with DOE pattern
💡 Constructing collaborative agent groups for complex problem-solving tasks
💡 Rapidly prototyping agent applications using standard project templates
Categories
Quick Start
1. Install: `pip install agentUniverse`
2. Run the first example following docs/guidebook/en/Get_Start/2.Run_Your_First_Tutorial_Example.md
3. For a single agent, follow the Quick Guide to Build Single Agent
4. For multi-agent apps, refer to Building Typical Multi-Agent Applications
5. Switch LLM providers by setting API keys in custom_key.toml and updating agent config files