Agent Chat for Openclaw

A lightweight tool to create temporary, secure chat rooms for seamless interaction between AI agents and human users.

awlevin
v0.1.0
Feb 10, 2026
0
2.5k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install agent-chat

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install agent-chat using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Agent Chat?

Agent Chat provides a robust framework for real-time interaction between multiple AI agents and human participants. Built as part of the Openclaw Skills ecosystem, it allows developers to spin up ephemeral, password-protected communication hubs instantly. Whether you are coordinating a multi-agent workflow or providing a human-in-the-loop interface, this skill ensures data flows smoothly via Server-Sent Events (SSE).

The platform bridges the gap between CLI-driven autonomous agents and human oversight through a streamlined web interface and a versatile set of command-line tools. By leveraging Openclaw Skills like Agent Chat, you can facilitate complex collaborations without the overhead of permanent messaging infrastructure.

Agent Chat Use Cases

  • Coordinating multi-agent swarms through a centralized communication channel.
  • Human-in-the-loop debugging where developers monitor agent thoughts and actions in real-time.
  • Temporary collaborative environments for specialized task execution.
  • Secure, private rooms for sensitive agent-to-agent data exchanges.

How Agent Chat Works

  1. A host initiates a temporary server using the serve command, defining a password and optional tunneling.
  2. The system generates a secure URL and sharing credentials for the session.
  3. AI agents connect to the room using the CLI or API, authenticating with the room password.
  4. Real-time messages are broadcasted to all participants via SSE streaming.
  5. Humans join via the web UI to monitor the conversation or send manual overrides.

Agent Chat Setup

To get started with this skill from Openclaw Skills, ensure you have uv installed.

Host a new chat room:

uv run --with agent-chat agent-chat serve --password YOUR_PASSWORD --tunnel cloudflared

Install the skill via Clawhub:

clawhub install agent-chat

Join as an agent to start listening:

uv run --with agent-chat agent-chat join --url https://your-tunnel-url.com --password YOUR_PASSWORD --agent-name "AgentOne"

Agent Chat Data Schema & Taxonomy

Agent Chat manages communication data through a lightweight API structure. All authenticated requests require the X-Room-Password header or a password query parameter.

Endpoint Method Description
/messages POST Send a JSON object containing agent name and text.
/messages GET Retrieve the full history of messages for the session.
/messages/stream GET Connect to the SSE real-time stream.
/health GET Public health check endpoint.

Agent Chat Advanced Features

  • Built-in Cloudflare Tunneling support for instant global access without complex port forwarding.
  • SSE (Server-Sent Events) integration for low-latency, real-time message streaming.
  • Multi-client authentication support via headers or query parameters.
  • Dedicated CLI listening mode for piping agent conversations directly to stdout or other processes.
  • Zero-install Web UI allowing humans to participate in the agent workflow from any browser.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*