Agent Definitions

AIGNE provides flexible AI agent creation methods. You can create various types of AI agents through simple configuration files, including chatbots, tool assistants, and complex intelligent systems.

Project Configuration File (aigne.yaml)#

aigne.yaml is the main configuration file for AIGNE projects, defining the overall project settings and the list of included agents.

Basic Structure#

Configuration Options#

• chat_model: Defines the default AI model configuration to use
• • provider: [Optional] Model provider, options include openai, claude, xai
  • name: Model name (e.g., gpt-4o-mini, gpt-4o, etc.)
  • temperature: Model output randomness (0.0-1.0). Lower values produce more deterministic output; higher values produce more diverse and creative output.
  • top_p: [Optional] Number of highest probability tokens to consider during sampling
  • frequency_penalty: [Optional] Reduces the probability of repeating tokens
  • presence_penalty: [Optional] Increases the probability of new tokens appearing
• agents: List of all agent configuration file paths included in the project

Agent Creation Methods#

AIGNE provides three different agent creation methods. You can choose the appropriate method based on your project requirements:

Creating AI Agents#

• Use YAML format agent definition files
• Suitable for quickly creating chatbots and simple AI assistants
• Simple configuration, easy to get started

Creating MCP Agents#

• Connect external resources through Model Context Protocol
• Support file system, database and other external tool integration
• Suitable for complex applications that need to access external data

Creating Agents through JS Functions#

• Use JavaScript functions to implement custom logic
• Support complete testing framework
• Suitable for advanced applications requiring complex business logic