pair-agent

Pair remote AI agents with a browser via HTTP for cross-agent control.

Updated Apr 21, 2026
One-click install
npx skills add https://github.com/scanbott/claude-skills --skill pair-agent-scanbott
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pair-agent
Source: https://github.com/scanbott/claude-skills/tree/main/gstack/pair-agent
Command: npx skills add https://github.com/scanbott/claude-skills --skill pair-agent-scanbott

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Pair a remote AI agent with your browser and securely share browser access so tasks can be performed across agents without manual handoffs.

Core Features & Use Cases

  • Remote browser orchestration: generate a setup key and provide clear instructions for the other agent to connect.
  • Cross-agent collaboration: supports OpenClaw, Hermes, Codex, Cursor, or any HTTP-capable agent with browser access.
  • Scoped access control: default read+write, admin on request; sessions are isolated per agent for safety.

Quick Start

To pair an agent, run the setup command to generate a key and share it with the remote agent; then the remote agent uses the key to connect and operate within the permitted scope.

Frequently Asked Questions about pair-agent

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I pair a remote AI agent with my browser for cross-agent control?

To pair a remote AI agent with your browser, generate a setup key and share it with the remote agent. The agent then uses this key to connect via HTTP requests and operate within its permitted scope.

Can I use Cursor or Codex to control my browser remotely?

Yes, you can use Cursor, Codex, OpenClaw, or Hermes to control your browser remotely. Any HTTP-capable agent can connect using the generated setup key to issue browser commands.

How does cross-agent browser collaboration handle session isolation and permissions?

Cross-agent browser collaboration provides each connected agent with its own isolated session tab. Permissions are scoped to read and write by default, with admin access available upon request for safety.

What is the best way to set up remote browser orchestration across multiple AI agents?

The best way to set up remote browser orchestration is running the setup command to generate a key. You then provide this key and connection instructions to the remote agent for immediate browser access.

Do I need to manually perform handoffs when orchestrating tasks across multiple AI agents?

No, you do not need manual handoffs when orchestrating tasks across multiple AI agents. Remote browser pairing enables agents to perform tasks collaboratively and securely without manual intervention.

Are there limitations on browser access when pairing remote AI agents?

Browser access limitations for paired remote AI agents are enforced through scoped permissions. Sessions are isolated per agent with default read and write access, restricting admin actions unless explicitly requested.