DevSnoop - Browser Control for AI Agents

DevSnoop gives coding agents a local bridge into your real Chrome browser. Instead of guessing from screenshots or editing blind, your agent can ask Chrome for compact page summaries, selectors, console errors, network failures, screenshots, DOM diffs, and basic page interactions. Use it when you want an agent to: - inspect what is actually rendered - find buttons, inputs, forms, and labels - click, fill, scroll, and hover in the current page - capture console errors and failed network requests - compare the page before and after a change - verify a fix in the browser before reporting back DevSnoop runs locally. The extension talks to a native host on your machine, and your coding agent calls that host over localhost. Page inspection results stay in your local agent workflow unless you choose to share them. Works with local coding agents and automation tools that can make HTTP requests to a local service. Setup has three parts: 1. Install this Chrome extension. 2. Install the DevSnoop native host from devsnoop.com/install. 3. Add the DevSnoop skill/instructions to your coding agent. The extension includes a free command preview so you can confirm the local bridge works before buying a license.