Script Planner Assistant▸ 
AI Apps 
AI News ANUS: An Open-Source AI Agent Framework for Automating Tasks via Natural Language and Multi-Agent Collaboration
小编 ·
ANUS: An Open-Source AI Agent Framework for Automating Tasks via Natural Language and Multi-Agent Collaboration
What is ANUS?
ANUS, which stands for "Autonomous Networked Utility System", is an open-source AI agent framework that can execute complex tasks through natural language instructions, support multi-agent collaboration, interact with web services, and handle document and code generation.
Main Features of ANUS
Task Automation and Multi-Agent Collaboration
ANUS supports single-agent and multi-agent modes and can dynamically switch based on task complexity. It allows users to create agents with different roles (such as researcher, analyst, writer, etc.) and collaborate through structured communication protocols.
Rich Tool Ecosystem
Web automation: supports browser operations, web scraping, form filling, and more.
Document processing: supports parsing and analysis of PDF and Office documents.
Code execution: provides a secure code execution environment and supports multiple programming languages.
Multimodal processing: supports the handling of images, audio, and video.
Flexible Model Integration
OpenAI API support: seamless integration with GPT-4 and newer models.
Open-source models: supports Llama, Mistral, and more.
Local deployment: run models locally to protect privacy and reduce costs.
Model switching: automatically selects the appropriate model based on task requirements.
Fallback mechanism: gracefully handles API issues by switching to alternative models.
Privacy and Security
Supports local execution of sensitive data processing to prevent data leakage. It also provides API key management, permission control, and audit logging features.
Extensibility
Plugin system: easily extend functionality through custom plugins.
Custom tools: create your own tools to extend agent capabilities.
Model adapters: add support for new AI models.
Middleware: insert custom processing steps into agent workflows.
Event hooks: respond to specific events in the agent lifecycle.
How to Install the ANUS AI Agent Framework?
Prepare the environment: ensure that the system has Python 3.11 or above installed, and that Git is configured.
Windows users can download Python from the official website.
macOS users can install it using the command brew install [email protected].
Linux users can install it using the command apt install python3.11.
1. Clone the repository:
git clone https://github.com/nikmcfly/ANUS.git cd ANUS
2. Create a virtual environment:
python -m venv venv source venv/bin/activate # Linux/macOS venv\Scripts\activate # Windows
3. Install dependencies:
pip install -r requirements.txt playwright install chromium # 安装浏览器内核[^31^]
4. Verify the installation: Check the version number to confirm successful installation.
Advantages of ANUS
Truly open source: accessible without invitation codes.
Hybrid architecture: combines the simplicity of single-agent setups with the power of multi-agent systems.
Flexible model support: supports OpenAI models, open-source models, or custom models.
Comprehensive tool ecosystem: includes web automation, document processing, code execution, and more.
Community-first design: encourages contribution and extension.
Transparent operations: clearly explains all agent behaviors and decisions.
Cross-platform: works across different operating systems and environments.
ANUS Use Cases
Simple task execution: such as querying information and generating reports.
Complex task collaboration: multiple agents working together to complete tasks such as research, analysis, and report writing.
Web automation operations: such as automatically filling out forms and scraping web data.
Code generation and execution: automatically generates and executes code.
GitHub repository address: https://github.com/nikmcfly/ANUS
