Aionda Browser MCP

Connect your AI assistant to Chrome for browser automation through a local MCP server. Requires Node.js and setup.

Connect your AI assistant to the Chrome tabs you already use. Aionda Browser MCP lets an MCP-compatible assistant read pages, capture screenshots and interact with websites through a local connection on your computer. Use it to test web interfaces, work through forms and automate browser tasks in your existing session. FEATURES • Read page text and interactive elements, with compact snapshots for smaller results. • Capture full-resolution or compressed screenshots of the visible tab. • Click, type, press keys, navigate and scroll. • List browser tabs, select a tab by URL or open a new one. • Use several AI assistants at the same time, each with its own tab. • Evaluate JavaScript expressions in an isolated page context. • Select local files for upload to a website. SETUP REQUIRED This extension is a companion to the free aionda-browser-mcp npm package. It is not a standalone AI assistant. You need Node.js 22 or newer and an MCP-compatible client on the same computer as Chrome. 1. Install the extension and open its Options page. 2. Install Node.js from https://nodejs.org/. 3. Add the server to your MCP client using command "npx" and arguments ["-y", "aionda-browser-mcp@0.2.1"]. The client downloads and starts the server automatically. Alternatively, install it with: npm install -g aionda-browser-mcp@0.2.1 4. Restart your MCP client. In the extension's Options, read the disclosure, check "Enable browser access" and save. 5. Open a website. The extension badge shows "on" when connected. Ask your assistant to check browser_status. Full instructions, configuration JSON and troubleshooting: https://github.com/AiondaDotCom/aionda-browser-mcp#installation LOCAL CONNECTION, YOUR CONTROL No Aionda account or cloud relay. The extension connects only to localhost. While access is enabled, each connected assistant follows your active tab until it selects or opens its own tab. Your MCP client receives tab information, requested page content, screenshots and action results; that client may send these to its AI provider. Enable access only for trusted clients and pages. Disable it in Options at any time. Chrome's Debugger API is used for coordinate clicks, JavaScript evaluation and file inputs. Chrome may display a debugging banner. Internal pages, Web Store pages and managed environments may restrict some actions. The extension and local server are free and open source; your chosen AI service may have its own costs. Privacy policy: https://aiondadotcom.github.io/aionda-browser-mcp/privacy.html