Skip to main content

Core Concepts

Understanding the fundamental concepts in Wegent.

Agents​

Agents are the core building blocks of Wegent. They are AI-powered entities that can perform tasks autonomously or interactively.

Types of Agents​

  1. Ghosts - Autonomous agents that work in the background
  2. Bots - Interactive agents that respond to commands
  3. Teams - Collections of agents working together

Skills​

Skills are capabilities that agents can use to perform tasks. They are:

  • Modular: Can be mixed and matched
  • Reusable: Shared across multiple agents
  • Extensible: Custom skills can be created

Tasks​

Tasks represent work to be done by agents. They include:

  • Task definition and requirements
  • Priority and scheduling information
  • Dependencies on other tasks

Models​

Models are the underlying AI/LLM configurations used by agents:

  • Supports multiple model providers
  • Configurable parameters (temperature, max tokens, etc.)
  • Model-specific optimizations

Shells​

Shells provide the execution environment for agents:

  • Command execution capabilities
  • File system access
  • Network operations

Configuration​

Everything in Wegent is configured through YAML files, making it easy to:

  • Version control your configurations
  • Share configurations across teams
  • Automate setup and deployment