intent_router

To route natural language to an AI agent system, you provide your goal as input. The router classifies the intent and proposes an ordered, auditable agent chain that remains pending until you explicitly approve it.

Intent routing for agent orchestration classifies natural language goals and maps them to an ordered sequence of agents. It proposes a fixed, auditable chain for user approval without executing any actions beforehand.

You set up an approval workflow by allowing the router to propose an ordered agent chain with rationale. The proposed chain remains pending and auditable until you review and explicitly approve the execution.

Yes, you can use natural language to coordinate planning and fixing agent workflows. The router supports understand, plan, fix, build, and audit intents, adapting the proposed agent chain to your current project state and available agents.

Yes, an intent router requires access to available agents and current task state. It uses this information to propose a relevant, ordered agent chain that adapts to your specific project environment and workflow coordination needs.