#785 · Primary category: AI Agents & Automation

python-sdk

agents computer-vision cua llms rpa test-automation vision-agents

Enable AI to control your desktop, mobile and HMI devices

Project last updated:08/28/26

GitHub Stars

551

Forks

61

Contributors

11

License

MIT

Why we included this project

UI test suites and RPA scripts tend to break the moment a button moves or a label changes, and this SDK attacks that problem by letting an AI agent see the screen instead of relying on brittle XPath or CSS selectors. You give it a high-level instruction like 'open a browser, search GitHub, and star a repo' and it works out the steps, or you can issue single-step commands such as click and type. It runs on Windows, Linux, macOS, and Android, and supports Anthropic, Gemini, and AskUI's own models out of the box, with a plugin path for custom providers if your team routes through Bedrock or Vertex. The MCP extension point also lets you add custom tools. Teams doing end-to-end testing, RPA, or building computer-use agents will find the multi-platform coverage and the ability to adapt to UI changes genuinely useful.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category