Agents

AI agents with instructions, a model, and tools.

Agents are the core building block of Aeontel. An agent combines a set of instructions, a language model, and one or more tools into an autonomous unit that can reason and act. When given a prompt, the agent reasons with its LLM, decides which tools to invoke, executes them, and returns structured output.

Each agent has a unique ID with the agt_ prefix. Agents can be run independently, composed into swarms for multi-agent delegation, or wired into workflows for multi-step pipelines.

Detailed documentation is coming soon. In the meantime, explore the API Reference or the platform directly.