# Ui.Vision MCP

Connect an AI agent to Ui.Vision for browser and desktop automation. Use it to run macros, inspect screenshots, or interact with on-screen controls through OCR.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 2087
- Author: A9T9
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/A9T9/RPA/tree/HEAD/mcp
- Updated: 2026-10-20
- HTML: https://affpapa.org/en/nevibecode/mcp/uivision-mcp

## What it does

The agent can create and run macros with the `uiv.*` API, then inspect logs and screenshots. Web pages can be controlled through the DOM or Chrome input; desktop apps and other screen content are handled visually with OCR and image search.

The `uivision-mcp-bridge` package relays requests to the Ui.Vision extension. Tools run in the side panel, so it needs to stay open; otherwise calls return a connection error. The server supports Claude Code, Codex, Cursor, GitHub Copilot, Windsurf, Cline, and Zed.

## Requirements
- Needs: Node.js (commands run via npx)
- Environment variables: `UIVISION_MCP_PORT` — keys and tokens: ask the human for the values, never commit them or put them in shared files.

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add --env UIVISION_MCP_PORT=<UIVISION_MCP_PORT> --transport stdio mcp -- npx -y uivision-mcp-bridge
```
Replace <…> with your own values

### Codex

**Platform CLI**
```bash
codex mcp add mcp --env UIVISION_MCP_PORT=<UIVISION_MCP_PORT> -- npx -y uivision-mcp-bridge
```
Replace <…> with your own values

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "mcp": {
            "command": "npx",
            "args": [
                "-y",
                "uivision-mcp-bridge"
            ],
            "env": {
                "UIVISION_MCP_PORT": "<UIVISION_MCP_PORT>"
            }
        }
    }
}
```
Replace <…> with your own values

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "mcp": {
            "type": "stdio",
            "command": "npx",
            "args": [
                "-y",
                "uivision-mcp-bridge"
            ],
            "env": {
                "UIVISION_MCP_PORT": "<UIVISION_MCP_PORT>"
            }
        }
    }
}
```
Replace <…> with your own values

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add -e UIVISION_MCP_PORT=<UIVISION_MCP_PORT> mcp npx -y uivision-mcp-bridge
```
Replace <…> with your own values

## When to use it
- control web pages through the DOM or Chrome input
- locate screen controls with OCR and image search
- run a macro and inspect its log
- interact with desktop apps using mouse and keyboard input

## Example prompts
- Use Ui.Vision OCR to find the button on screen and click it.
- Run the `checkout` macro and inspect its execution log.
- Take a screenshot and describe what is open in the browser.

## Similar
- [vision](https://affpapa.org/en/nevibecode/skills/vision.md) — skill · 13400 installs
- [qianwen-vision](https://affpapa.org/en/nevibecode/skills/qianwen-vision.md) — skill · 5700 installs
- [vision-framework](https://affpapa.org/en/nevibecode/skills/vision-framework.md) — skill · 3800 installs
- [fal-vision](https://affpapa.org/en/nevibecode/skills/fal-vision.md) — skill · 2700 installs
- [vision-analysis](https://affpapa.org/en/nevibecode/skills/vision-analysis.md) — skill · 2700 installs
- [axiom-vision](https://affpapa.org/en/nevibecode/skills/axiom-vision.md) — skill · 1800 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md