@wsl8297: Found an interesting open-source automation AI tool on GitHub: Terminator. Instead of relying on OCR to 'see' the screen, it directly 'reads' the application structure like parsing HTML, making it faster and more accurate at controlling various software on your computer. Compared to traditional OCR automation tools...
Summary
Introduces an open-source automation AI tool called Terminator, which directly reads application structures instead of relying on OCR, enabling faster and more precise control of computer software. It supports Windows and partial macOS features.
View Cached Full Text
Cached at: 05/17/26, 03:28 AM
I found an interesting open-source automation AI tool on GitHub: Terminator. Instead of relying on OCR to “see” the screen, it directly “reads” the application structure—like parsing HTML in a web page—making it faster and more accurate at controlling various software on your computer. Compared to traditional OCR-based automation tools, its advantages are clear:
- Faster: More than just a marginal speed boost
- Works even when obscured: Can still interact with windows that are in the background or covered by other windows
- More stable: Small UI changes are less likely to break your workflows
Project link: https://github.com/mediar-ai/terminator
Currently supports Windows primarily; macOS only has partial functionality. If you want to automate tasks on your desktop and understand the underlying implementation logic, this project is worth checking out.
mediar-ai/terminator
Source: https://github.com/mediar-ai/terminator
Demo
Computer Use MCP that controls your entire desktop
Give AI assistants (Claude, Cursor, VS Code, etc.) the ability to control your desktop and automate tasks across any application.
Claude Code (one-liner):
claude mcp add terminator "npx -y terminator-mcp-agent@latest"
Other clients (Cursor, VS Code, Windsurf, etc.):
Add to your MCP config file:
{
"mcpServers": {
"terminator-mcp-agent": {
"command": "npx",
"args": ["-y", "terminator-mcp-agent@latest"],
"env": {
"LOG_LEVEL": "info",
"RUST_BACKTRACE": "1"
}
}
}
}
See the MCP Agent README (https://github.com/mediar-ai/terminator/tree/main/terminator-mcp-agent) for detailed setup instructions.
Why Terminator MCP?
- Uses your browser session - no need to relogin, keeps all your cookies and auth
- Doesn’t take over your cursor or keyboard - runs in the background without interrupting your work
- Works across all dimensions - pixels, DOM, and Accessibility tree for maximum reliability
Use Cases
- Create a new instance on GCP, connect to it using CLI
- Check logs on Vercel to find most common errors
- Test my app new features based on recent commits
What’s new
- 01/09/26 - Mediar IDE (Cursor for Windows automation) is in public access - download now (https://cdn.crabnebula.app/download/mediar/mediar/latest/platform/windows-x86_64)
- 10/30 Public alpha is live - Cursor for Windows automation (https://www.mediar.ai)
- 09/26 Terminator was on Cohere Labs podcast (https://www.youtube.com/watch?v=cfQxlk8KNmY), also check the slides (https://092025-cohere.mediar.ai/)
- 08/25 Big release — NodeJS SDK in YAML workflows, run JS in browser, OS event recording → YAML generation in MCP, and more
- 08/25 we raised $2.8m to give AI hands to every desktop (https://x.com/louis030195/status/1948745185178914929)
Why Terminator
For Developers
- Create automations that work across any desktop app or browser
- Runs 100x faster than ChatGPT Agents, Claude, Perplexity Comet, BrowserBase, BrowserUse (deterministic, CPU speed, with AI recovery)
-
95% success rate unlike most computer use overhyped products
- MIT-licensed — fork it, ship it, no lock-in
We achieve this by pre-training workflows as deterministic code, and calling AI only when recovery is needed.
For Teams
Our public beta workflow builder (https://www.mediar.ai) + managed hosting:
- Record, map your processes, and implement the workflow without technical skills
- Deploy AI to execute them at >95% success rate without managing hundreds of Windows VMs
- Kill repetitive work without legacy RPA complexity, implementation and maintenance cost
Feature Support
Terminator currently supports Windows only. macOS and Linux are not supported.
| Feature | Windows | macOS | Linux | Notes |
|---|---|---|---|---|
| Core Automation | ||||
| Element Locators | Yes | No | No | Find elements by name, role, window, etc. |
UI Actions (click, type) | Yes | No | No | Core interactions with UI elements. |
| Application Management | Yes | No | No | Launch, list, and manage applications. |
| Window Management | Yes | No | No | Get active window, list windows. |
| Advanced Features | ||||
| Browser Automation | Yes | No | No | Chrome extension enables browser control. |
| Workflow Recording | Yes | No | No | Record human workflows for deterministic automation. |
| Monitor Management | Yes | No | No | Multi-display support. |
| Screen & Element Capture | Yes | No | No | Take screenshots of displays or elements. |
| Libraries | ||||
Python (terminator.py) | Partial | No | No | pip install terminator |
TypeScript (@mediar-ai/terminator) | Yes | No | No | npm i @mediar-ai/terminator |
Workflow (@mediar-ai/workflow) | Yes | No | No | npm i @mediar-ai/workflow |
CLI (@mediar-ai/cli) | Yes | No | No | npm i @mediar-ai/cli |
KV (@mediar-ai/kv) | Yes | No | No | npm i @mediar-ai/kv |
MCP (terminator-mcp-agent) | Yes | No | No | npx -y terminator-mcp-agent --add-to-app [app] |
Rust (terminator-rs) | Yes | No | No | cargo add terminator-rs |
Legend:
- Yes: Supported - The feature is stable and well-tested.
- Partial: Partial / Experimental - The feature is in development and may have limitations.
- No: Not available on this platform.
How to Inspect Accessibility Elements (like name:Seven)
To create reliable selectors (e.g. name:Seven, role:Button, window:Calculator), you need to inspect the Windows Accessibility Tree:
Windows
- Tool: Accessibility Insights for Windows (https://accessibilityinsights.io/downloads/)
- Alt: Inspect.exe (https://learn.microsoft.com/en-us/windows/win32/winauto/inspect-objects) (comes with Windows SDK)
- Usage: Open the app you want to inspect → launch Accessibility Insights → hover or use keyboard navigation to explore the UI tree (Name, Role, ControlType, AutomationId).
These tools show you the
Name,Role,ControlType, and other metadata used in Terminator selectors.
Platform Support
| Platform | CLI | MCP Agent | Automation | Installation Method |
|---|---|---|---|---|
| Windows | Yes | Yes | Yes | npm/bunx |
Note: Terminator currently supports Windows only. macOS and Linux support is not available.
Troubleshooting
For detailed troubleshooting, debugging, and MCP server logs, send us a message (https://www.mediar.ai/).
Contributing
Contributions are welcome! Please feel free to submit issues and pull requests. many parts are experimental, and help is appreciated.
Similar Articles
@VincentLogic: Found an incredible open-source desktop AI tool from ByteDance! UI-TARS Desktop, with 31k stars, truly lives up to the hype. It can actually understand your screen and automate computer operations for you. Just tell it "Enable auto-save in VS Code and set the delay to 500ms", and it will automatically: -…
ByteDance's open-source desktop AI automation tool, UI-TARS Desktop, supports local execution and screen visual understanding. It can autonomously control your computer to handle daily tasks through natural language commands.
@GitHub_Daily: Microsoft recently open-sourced an AI terminal: Intelligent Terminal, based on Windows Terminal, with an AI assistant built into the terminal. It can automatically detect command-line output, and when errors occur, send the context to AI for analysis with one click, without manual copy-paste, and can even directly help execute...
Microsoft open-sourced Intelligent Terminal, an AI-powered terminal based on Windows Terminal that integrates AI assistants to help with command-line errors and tasks.
@JiweiYuan: Recently built a Mac-native AI reading tool called OakReader. While reading PDF papers/eBooks or browsing the web, when you're confused, just press a button to open the sidebar, select a passage of text, and you can discuss it with your favorite AI model without ever having to switch to ChatGPT / Claude / …
OakReader is a Mac-native AI reading tool that supports reading PDF papers/eBooks and web pages. It comes with built-in sidebar AI chat, note-taking, translation, browser, and reference management features, and can be configured with multiple AI models.
@billtheinvestor: The 7MB AI-native terminal is here! Terax perfectly integrates a terminal, code editor, Git, and a powerful AI agent, and also supports local models. Lightweight beyond belief, blazingly fast — the developer's dream machine! GitHub link:
Terax is a lightweight open-source AI-native terminal and developer workspace that combines a terminal, code editor, Git, and AI agent in a 7MB package, supporting local models and various AI providers.
@CycleDecoded: Programmers and terminal freebie hunters, look here — someone actually built this. In the past, writing code meant opening a bunch of windows, switching back and forth between the black Terminal, VSCode, browser, and ChatGPT until your hands were worn out. The open-source programmer project Wave Terminal (waveterm) directly combines the command line, ...
Introducing the open-source AI terminal Wave Terminal, which integrates the command line, editor, browser, and AI into one terminal, with support for inline previews, persistent SSH sessions, and multi-model access. It has received 22k+ Stars on GitHub.