agent-eyes

mohanli-eng/agent-eyes
★ 0 stars TypeScript 🤖 AI/LLM Updated today
Give Claude Code eyes to see the browser runtime — an MCP server + workflow skill for AI coding agents.
View on GitHub →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "agent-eyes": {
      "command": "npx",
      "args": [
        "-y",
        "mohanli-eng/agent-eyes"
      ]
    }
  }
}

README Excerpt

> A modular harness for AI coding agents — giving Claude Code eyes to see the browser runtime. --- Modern AI coding agents (Claude Code, Cursor, Codex CLI) can write code, run tests, and ship implementations. But they share a structural blind spot: **once code is deployed, they can't see what actually happens in the browser at runtime**. They write a React component, the unit tests pass, the build succeeds — and then a `useEffect` infinite loop floods the console, an API call returns 500, or a C

Tools (3)

getDOMSnapshotgetNetworkRequestsgetPerfMetrics

Topics

agent-toolingai-agentsbrowser-automationclaude-codedeveloper-toolsmcpplaywrightruntime-observation