About Agentrq
AgentRQ is a human in the loop task manager for AI agents. It gives autonomous coding agents such as Claude Code a real time channel to the person supervising them: agents assign tasks and questions to the human, the human assigns tasks back, and both sides keep working in parallel.
It is aimed at developers running one or more agents who want to approve decisions and answer questions from anywhere, including a phone. Each workspace is isolated to a single agent with its own persona, mission and dedicated MCP endpoint, and a supervisor MCP server lets one agent hand tasks to others. A conversational dashboard shows every task, thread and decision as a list or Kanban board. Claude Code gets native channel push notifications, and an open source gateway connects agents that speak the Agent Client Protocol, including Gemini CLI, Codex CLI, Cursor, GitHub Copilot and OpenCode. Other features include events that create the next task when one finishes, visual workflows that chain workspaces, one time and recurring scheduled tasks, a searchable history of every tool call, and in browser models for task titles and speech to text dictation.
AgentRQ is free during its beta, with no credit card required. It is also open source under the Apache 2.0 license and can be self hosted. No paid plans have been announced.
The product is built on MCP notification channels, so the depth of integration varies by agent, with Claude Code receiving the most direct support.