Convex Agents Human Agents

What problem does it solve?

Humans are needed in automated workflows to provide context, guidance, and approvals when AI alone cannot determine the best course of action. This skill enables human-in-the-loop interactions within Convex agent threads, supporting escalation, validation, and collaborative decision-making.

Core Features & Use Cases

• Human-in-the-loop collaboration: route, store, and respond to messages from humans and AI agents in the same thread.
• Escalation and approvals: support escalation to humans and approval gates before sending high-stakes responses.
• Assignment and routing: track who should respond next and route threads accordingly for mixed AI/human workflows.
• Tool-based interventions: trigger human involvement through tools when AI detects the need for human input.

Quick Start

Route a thread to a human agent for review and response when AI cannot decide.