empir3hq

Empir3 Bridge — AI skill for Claude Code

AI community

Local-first MCP bridge — orchestrate the AI CLIs you already pay for (Claude, Codex, Gemini, Grok) and drive a real browser + desktop, all from one agent.

How to install Empir3 Bridge

This entry records only its repository, not the path inside it, so there is no exact command to give. Open empir3hq/empir3-bridge and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Empir3 Bridge does

Local-first MCP bridge — orchestrate the AI CLIs you already pay for (Claude, Codex, Gemini, Grok) and drive a real browser + desktop, all from one agent.

Alternatives in AI

Botmux — Bridge Feishu/Lark to AI coding CLIs — Claude Code, Codex, Gemini, OpenCode… every DM, group or topic spawns i 1.2k ★
Agor — Orchestrate Claude Code, Codex, and Gemini sessions on a multiplayer canvas 1.2k ★
UltraCode Shim — Give Claude Code's ultracode mode to ANY model you already pay for 422 ★

README

Empir3 Bridge

Your AI subscriptions, finally working as a team.

https://github.com/user-attachments/assets/1db136ff-c704-4232-9444-1453a3ea2fa7

**▶ 60 seconds** — browser use, computer use, and multi-model orchestration: your subscriptions, your machine, no paywall.

Codex, Grok, Gemini, and Claude — the seats you're **already paying for** — wired into a single conversation where one agent quarterbacks the rest. Have Codex build it, Grok draft it, Gemini check it, Claude tie it all together, and pull every answer back into one thread.

**No API keys handed out. No per-token meter running.** Just the flat-rate CLIs you're already signed into — suddenly collaborating instead of sitting in four lonely terminals talking to nobody.

Then the team acts where chat can't: a real browser and your live desktop, driven by **pixel-accurate, OS-level clicks** that map page coordinates to physical screen pixels — landing on the canvas, trusted-event, and React-Native-Web UIs that synthetic automation can't touch.

**Three ways to drive it — MCP, CLI, or WebSocket.** MCP for the agents you talk to, CLI for your terminal and scripts, or WebSocket via [empir3.com](https://empir3.com) for paired remote control. Same engine behind all three — pick the door that fits.

![Terminal demo — one prompt ("Have Codex scaffold the API, Gemini review the diff, then open the dashboard in a real browser and screenshot it") fans out to Codex on your OpenAI seat, Gemini on your Google seat, and the bridge's own Chrome, with every result reported back into one thread — no API keys, no per-token meter](assets/demo-orchestration.svg)

One prompt, the whole team — Codex on your OpenAI seat, Gemini on your Google seat, the browser on this machine, all in one thread. ([how it's wired ↓](#architecture))

![One agent in any MCP client delegates through the local Empir3 Bridge to your Codex, Grok, Gemini, and Claude seats — plus a dedicated Chrome and a scoped region of your d