The agent responsible for executing specific tasks within an agentic AI framework is typically called a worker (or task/executor) agent.

Quick Scoop

In an agentic AI setup, different agents have different roles, and the worker agent is the one that actually “does the work.”

  • It takes concrete instructions or subtasks from higher-level agents (like planners or orchestrators).
  • It then performs specific actions such as running a tool, processing data, updating a system, or generating a draft output.
  • Unlike orchestrator or planner agents, it does not decide what should be done overall, only how to carry out the assigned task in context.

Mini breakdown of roles

  • Orchestrator / reasoning agent: Breaks a big goal into subtasks, routes them to the right agents, and manages the workflow.
  • Planner agent: Designs the sequence or strategy for solving the overall problem.
  • Worker / task / executor agent: Executes the specific, well-defined tasks it is given; this is the answer to your question.
  • Monitoring / observer agent: Tracks progress, checks outcomes, and can raise alerts if something goes off track.

In short: planners and orchestrators think, worker agents act.