A system made up of multiple AI agents, each with a distinct role, that work together to complete a complex task. In a testing context, a multi-agent system might include separate agents for exploring an application, generating tests, executing them, detecting defects, and orchestrating the overall workflow.
The agents can communicate with each other, raise issues between themselves, and make collective decisions, such as whether a test failure represents a genuine bug or a poorly written test. Multi-agent systems allow tasks to be parallelised and handled at a scale that would not be feasible with a single agent or a human team alone.
The agents can communicate with each other, raise issues between themselves, and make collective decisions, such as whether a test failure represents a genuine bug or a poorly written test. Multi-agent systems allow tasks to be parallelised and handled at a scale that would not be feasible with a single agent or a human team alone.
