Agent Peer via Tailscale for Openclaw

Enable direct, encrypted communication between Openclaw agents on different networks using Tailscale VPN without complex port forwarding.

fuzzyb33s
v1.0.0
Apr 11, 2026
0
779
0

Install & Download

1. ClawHub CLI

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

npx clawhub@latest install agent-peer-tailscale

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-peer-tailscale 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 Peer via Tailscale?

The Agent Peer via Tailscale skill provides a robust framework for connecting two Openclaw agents running on separate physical machines. By leveraging Tailscale VPN, this skill bypasses the need for public IPs, complex port forwarding, or intermediary relay servers. It creates a secure, private network where agents can communicate as if they were on the same local area network.

This architecture allows for seamless use of the sessions_send function across different hosts. It is an essential addition for developers building decentralized Openclaw Skills networks, allowing agents to share session context, provide real-time tips, and delegate complex tasks to specialized peers safely and efficiently.

Agent Peer via Tailscale Use Cases

  • Collaborative coding projects where two agents on different machines need to synchronize progress.
  • Mentor-peer relationships where a senior agent provides real-time insights and code reviews to another.
  • Decentralized agent networks that require cross-machine task delegation.
  • Sharing session context between home and office machines without exposing services to the public internet.

How Agent Peer via Tailscale Works

  1. Both machines install Tailscale and join the same private mesh network to receive a stable internal IP.
  2. The Openclaw gateway on each machine is configured to bind to the Tailscale interface (0.0.0.0) to accept peer traffic.
  3. Users exchange Tailscale IPs and gateway tokens to establish mutual trust.
  4. A peer configuration file is created to map agent IDs to their respective Tailscale gateway URLs.
  5. Agents utilize the sessions_send command with the specific peer gatewayUrl to transmit messages and data securely.

Agent Peer via Tailscale Setup

1. Install Tailscale

On your primary machine, install and authenticate:

tailscale up --accept-routes
tailscale ip -4

2. Configure Openclaw Gateway

Modify your Openclaw configuration to allow external peer connections over the VPN:

{
  "gateway": {
    "bind": "0.0.0.0",
    "port": 8080
  }
}

Restart the gateway to apply changes:

openclaw gateway restart

3. Verify Connection

Test connectivity to the peer machine:

curl http://<peer-tailscale-ip>:8080/health --connect-timeout 5

Agent Peer via Tailscale Data Schema & Taxonomy

The skill organizes peer information through a dedicated configuration file and supporting reference documents:

File Description
peer-agent/peer-config.md Stores Tailscale IPs, Gateway URLs, and Auth Tokens for both agents.
references/tailscale-setup.md Detailed network setup and auth key management documentation.
references/peer-communication.md Specifies message formats and session management protocols.
scripts/peer_config.py An interactive tool for generating the required peer configurations.

Agent Peer via Tailscale Advanced Features

  • Direct task delegation: Pass complex workloads to a peer agent with specialized tools.
  • Morning Handoff patterns: Automate status updates and blocker synchronization between agents.
  • Cross-machine code review: Request peer agents to analyze and suggest improvements to local code snippets.
  • Secure session targeting: Use gateway.access to strictly control which sessions a peer can interact with.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*