ralplan

Multi-agent consensus planning is automated by coordinating Planner, Architect, and Critic agents that iterate until a plan reaches approval. This workflow handles multiple feedback cycles and architectural checks to ensure plan quality.

Iterative consensus planning is a review process where specialized agents draft, architecturally review, and critique a plan until agreement is reached. It is needed for multi-step tasks requiring refinement and architectural checks before execution.

To run a consensus planning cycle, trigger the workflow and the system will enforce user feedback prompts at review and approval stages during each iteration. This ensures alignment with your goals before delegating execution.

Consensus planning requires the Plan skill to be running in consensus mode. It acts as a compact alias to trigger this specific workflow and delegate safe execution handoff upon final approval.

Multi-agent review supports tasks needing architectural checks by deploying an Architect agent to review the initial plan drafted by the Planner. The Critic then evaluates overall quality until consensus is achieved.

Handling multiple feedback cycles in plan refinement is best achieved through an iterative consensus workflow that coordinates specialized agents. It enforces user feedback prompts during each iteration until the plan reaches final approval.