MCP serverCautionBrowser automation & web dataSome setup Verified live

RPA

This MCP server enables AI agents to automate browser and desktop tasks using macros, JavaScript, OCR, and image recognition.

Browser and desktop automation with an MCP server for AI assistants. Record macros or write JavaScript to automate websites, extract data, and control desktop apps with OCR and image recognition.

2.1k stars415License: No licenseUpdated 13 days ago0 opens

Preview

▍

What you can do with it

Automate data extraction from websites.
Control desktop applications for repetitive tasks.
Test website functionality by replaying interactions.

Verified live · 25 tools

We started this MCP server in an isolated sandbox and it exposed these tools:

bridge_status
Reports whether the Ui.Vision browser extension is currently connected to this bridge. Call this first if other tools fail — it tells you whether the extension side panel is open and the bridge is ena
run_status
What the macro runner is doing RIGHT NOW: whether a run is in progress, which macro the editor holds, and the freshest log lines. Unlike every other tool, this ANSWERS WHILE run_macro is still executi
select_browser
Switches which connected browser the other tools talk to. Several browsers (or several instances of one browser) can hold a bridge connection at the same time; bridge_status lists their labels. Pass a
open_panel
Opens the Ui.Vision panel when the extension is NOT connected: signals the browser over a background wake channel (extension 10.0.170+) to open the Ui.Vision app in a tab. This is the ONLY way the pan
list_macros
Lists all macros stored in Ui.Vision (name and id). Use open_macro to load one into the editor before reading or running it.
open_macro
Opens the macro with the given name in the Ui.Vision editor and returns its JSON. Subsequent get_macro/set_macro/run_macro calls operate on it. Note: unsaved editor changes may be discarded.
get_authoring_guide
Returns the Ui.Vision macro authoring guide: the uiv.* JavaScript API for JS script macros (finders, actions, OCR, vision), the classic command table, locator rules and best practices. CALL THIS ONCE
get_macro
Returns the macro currently loaded in the Ui.Vision editor, as Ui.Vision JSON. A JS script macro comes back with a "Script" field (its program) instead of Commands.
send_chat
Sends a message to the REAL in-panel AI Chat — the configured chat model runs it with its own agent tools, exactly as if a user had typed it. Returns immediately; a chat run takes minutes, so poll get
get_chat
Returns the in-panel AI Chat transcript and whether a run is in progress (running:true/false). Poll this after send_chat; the transcript is text-only (inline images are omitted).
reload_extension
Reloads the Ui.Vision extension from disk (same as the chrome://extensions reload button) — the way to load a freshly built version without a human click. The bridge connection dies with the panel; on
run_selftest
Runs every self-test of Settings > Desktop Automation in one go: native host connection, WASM image-search engine, desktop screenshot, real mouse input delivery, home-directory read/write, visual comp
get_ai_settings
Reports the ACTIVE AI provider/model as the chat will resolve it, which providers have keys stored (booleans only — keys never cross the bridge), the per-provider model settings, computer-vision scope
set_ai_settings
Switches AI settings for independent runs (e.g. benchmark sweeps): provider (uivision | openrouter | anthropic | local), uivision_tier (free | pro), openrouter_model / anthropic_model / local_model /
set_macro
Apply changes to the macro in the Ui.Vision editor — pass the complete Ui.Vision JSON. PREFER the JS script form {"Name": "name", "Script": "<JavaScript using the uiv.* API>"} over the classic Command

npx -y uivision-mcp-bridge · 06/10/2026

Safety check

Caution

Usable, but read the notes first.

●No clear license: check with the author before using it in a commercial product.

Automated scan of the repo files. Always review code you run.

From the README

Ui.Vision - Browser and Desktop Automation

Automate websites and desktop applications with macros, JavaScript, or an AI assistant. Ui.Vision combines browser automation, OCR and image recognition, with an MCP server that lets AI agents create, edit and run reusable macros in your browser.

Install the extension · Connect an AI assistant with MCP · API reference · User forum

What can you automate?

Browser tasks: fill forms, extract web data, download reports and repeat workflows in your existing browser session.
Website testing: record and replay interactions, use Selenium IDE commands, and run checks with different input data.
Visual tasks: find on-screen text with OCR and locate controls by image when HTML selectors are unavailable.
Desktop workflows: automate applications and remote desktop interfaces using visual recognition and mouse and keyboard input. Desktop automation requires the additional Ui.Vision desktop components (XModules).
AI-assisted automation: ask the built-in AI assistant or an external MCP client to write and run macros, then inspect the resulting script, logs and screenshots.

Use the macro recorder and command table, write JavaScript with the uiv.* API, or let an AI assistant author the macro. The resulting automation can be saved and run again.

Get started

Install the browser extension; you do not need to build this repository:

Open Ui.Vision and record a browser task, or start with a macro from the extension's demos. For desktop input and additional native capabilities, see the XModules installation guide.

The browser extension is open source and free for personal and commercial use. See the Ui.Vision website for desktop components and available editions.

Connect an AI assistant with MCP

Ui.Vision provides a Model Context Protocol (MCP) server through the uivision-mcp-bridge package. It connects MCP clients such as Claude Code, Claude Desktop and Cursor to the Ui.Vision browser extension.

An agent can inspect a page, create or edit a macro, run it, and use logs and screenshots to check the result. The automation executes through Ui.Vision in your browser.

Start with the MCP setup guide. The installer command is:

npx uivision-mcp-bridge --setup

The installer registers Ui.Vision with supported MCP clients found on your machine. Then enable the MCP bridge in Ui.Vision → Settings → AI, complete the connection steps in the guide, and restart your MCP client. Node.js and npm are required to run the installer.

For a first task, ask your connected assistant:

Use Ui.Vision to open https://example.com, read the page heading, and save a reusable macro for this task.

See mcp/README.md for bridge source, configuration and troubleshooting.

JavaScript automation and documentation for AI assistants

The uiv.* JavaScript API covers page elements, visual matching, OCR, browser and desktop input, screenshots, tabs, CSV files and downloads. Ui.Vision macros execute sequentially: write API calls without async or await.

When asking an AI assistant to write a macro, give it the API reference below so it uses Ui.Vision's supported methods and runtime conventions.

Ui.Vision JavaScript API and AI authoring guide: API syntax and automation recipes used by the built-in assistant.
Plain-text reference for AI assistants: documentation that can be supplied directly as context.
Classic command-to-JavaScript mapping: uiv.* equivalents and the uiv.run('command', 'target', 'value') bridge for classic commands.
Selenium IDE command reference: documentation for command-table macros.

Help and support

Ask questions and share automation examples in the Ui.Vision user forum, where users, support staff and developers participate.

Build from source

Building the extension is not required if you "only" want to use it.

You can install UI.Vision directly from the Chrome, Edge or Firefox stores, which is the easiest and the recommended way of using the Ui.Vision browser extension. Older versions can be found in the RPA software archive.

The information below is only required and intended for developers:

The project uses Node V20.11.1 and NPM V10.2.4

If you have any questions, please contact us at TEAM AT UI.VISION - Thanks!

Build the extension bundle

npm i -f

npm run build

npm run build-ff

npm run build creates the Chrome/Edge build in dist, npm run build-ff creates the Firefox build in dist_ff.

Develop

npm i -f

npm start

Use npm run start-ff for the Firefox variant. Both run webpack in watch mode, so the bundles in dist (Chrome/Edge) and dist_ff (Firefox) are rebuilt on every change.

Once done, the ready-to-use extension code appears in the /dist directory (Chrome, Edge) or /dist_ff directory (Firefox). Load it via chrome://extensions → "Load unpacked" (Chrome/Edge) or about:debugging → "Load Temporary Add-on" (Firefox).

Repository layout

src/ - the extension source (React UI, side panel, macro player, commands)
extension/ - static extension assets and manifest.json (Manifest V3)
mcp/ - the Ui.Vision MCP bridge, which lets Claude Code an

View source on GitHub