intelligent-routing

Automatic agent routing removes the need to explicitly name agents by analyzing request intent, domains, and complexity to pick the best specialist for tasks like authentication, UI, API, or database requests.

You route a security review request by submitting the task directly, such as asking to review authentication flows for vulnerabilities. The system classifies the intent and automatically invokes the security review specialist.

Yes, you can override automatic routing. The system supports explicit override handling, meaning it respects direct user agent mentions, such as using @backend-specialist, when provided in the request.

The best way to delegate cross-domain workflow automation tasks is using a deterministic decision-tree classification and agent selection matrix. It routes complex requests to an orchestrator for coordinated multi-agent execution.

Yes, intent classification supports deployment and performance testing requests. The agent selection matrix maps common intents, including DevOps deployment and performance optimization, to their respective specialist agents.

You should not use automatic task delegation when you need precise control over a specific specialist. In these cases, bypass intent classification and explicitly mention the desired agent to trigger the override handling.