goap-agent

GOAP planning decomposes complex tasks into manageable steps, maps dependencies between them, and creates executable plans that coordinate multiple specialized agents. This enables systematic task breakdown and dependency tracking before execution begins.

Multi-agent coordination supports parallel, sequential, swarm, and hybrid execution strategies. Choose parallel for independent tasks, sequential for ordered dependencies, swarm for iterative refinement, or hybrid to combine strategies optimally for your task structure.

Quality gates enforce validation checkpoints between task stages, ensuring each agent's output meets standards before downstream agents proceed. This prevents cascading failures and maintains output quality across the execution pipeline.

Yes, GOAP planning applies to software engineering tasks requiring decomposition, dependency mapping, and coordinated execution—such as build orchestration, testing workflows, deployment pipelines, and cross-functional project coordination.

The planning process generates executable plans stored in a dedicated plans/ directory, capturing task decomposition, dependency relationships, and execution strategies. These artifacts document the coordination strategy and enable reproducible execution.

Yes, GOAP planning enables invocation of other Skills and Tasks as part of the coordinated execution. This allows modular composition where specialized agents—implemented as Skills or Tasks—handle specific subtasks within the overall plan.