ralplan

Multi-agent consensus planning coordinates a Planner, Architect, and Critic in an iterative loop to validate decisions. This mechanism produces a robust plan by applying thorough verification and risk-aware checks to tasks like software changes or process redesigns.

Use deliberate planning mode for high-risk tasks that require thorough verification. It triggers pre-mortems and expanded test planning, ensuring your workflow anticipates potential failures and validates decisions before finalizing policy drafts or software changes.

Start a consensus planning workflow by describing your task to initiate the iterative loop. The Planner, Architect, and Critic agents will process your input, collect feedback through interactive prompts, and deliver a validated plan with concrete verification steps.

Yes, interactive prompts guide you at key decision points to tailor the planning cycle. This feature allows you to provide specific input during the iterative loop, ensuring the final plan aligns with your risk-management and collaboration requirements.

The final output of a multi-agent planning workflow is a validated plan. It includes concrete drivers, alternatives considered during the iterative loop, and specific verification steps to ensure your task meets risk-aware decision-making standards.

Yes, multi-agent consensus planning applies to process redesigns and policy drafts. The Planner, Architect, and Critic loop provides thorough verification and risk-aware decision-making for any complex task requiring a validated and robust plan.