Human-in-the-Loop Task Manager for AI Agents
Docs / ACP-Compatible Agents
Agent Directory

ACP-Compatible Agents

Any agent that speaks the Agent Client Protocol (ACP) can connect to AgentRQ through one open-source gateway — no custom integration code per agent. Below is every ACP-compatible agent we're tracking, with a direct guide for each.

What Does "ACP-Compatible" Mean?

The Agent Client Protocol is an open standard — originally developed by the Zed team — that lets any coding agent talk to any client over a standard JSON-RPC interface, the same way the Language Server Protocol standardized how editors talk to language tooling. An agent that's "ACP-compatible" can run as a subprocess and speak that same protocol, regardless of which client is driving it.

That matters for AgentRQ because it means we don't need to write a custom integration for every agent on this page. AgentRQ's open-source ACP Gateway implements the client side of ACP once — any agent below plugs straight in.

Why Connect Your Agent to AgentRQ?

Real-Time Task Board

Every task your agent works on appears on your AgentRQ dashboard instantly — no polling, no black box.

Approve From Anywhere

Approve or deny risky commands from your phone, another monitor, or Slack — without being at the terminal.

One Gateway, Any Agent

Switch agents without switching your human-in-the-loop workflow — the same ACP Gateway works for all of them.

Full Audit Trail

Every tool call, reply, and decision is recorded and searchable in your workspace's tool call history.

All Agents

Don't see your agent, or building your own? Any agent that implements ACP works the same way — follow the generic ACP Gateway guide. Running Claude Code directly? It connects natively via MCP — see the Claude Code guide instead.

Setup Complete?
Learn how to manage your workspace next.