Skip to content

Planning

Before any task runs, AIgency's planner turns your objective into a bounded set of tasks, dependencies, and acceptance criteria.

AgentDock Pro

Planning is part of AIgency, available only in AgentDock Pro.

Creating a run doesn't mean writing out tasks yourself — see Creating a run for the objective/constraints/completion-criteria fields you actually fill in. The planner is what turns that into an executable dependency graph.

What the planner produces

  • A set of tasks, each scoped to a bounded unit of work rather than the whole objective at once.
  • Dependencies between tasks, where one task's output is needed before another can start.
  • An initial agent assignment per task, using task routing's capability scoring.
  • Draft acceptance criteria, where applicable, that later feed verification.

Planner output is a proposal, not a commitment

A run stays in draft status while you review the planned tasks, and moves to ready once you're satisfied — see Creating a run for the full run-status lifecycle. You can edit, remove, or reassign any planner-proposed task before the run starts, and acceptance criteria the planner proposes are treated the same untrusted-until-you-approve-them way as everything else it drafts.