Back to catalog
TagUI
Apache-2.0TagUI is an open-source RPA tool replacing costly enterprise suites like UiPath with plain-text, command-line flows for web and desktop automation.
Open-source alternative to — compare it against:
Overview
TagUI transpiles plain text flows into executable JavaScript to automate user actions across web browsers and desktop software. It runs locally from the command line on Linux, macOS, and Windows without agent daemons or subscription servers. By combining Chrome DevTools Protocol for web selectors with SikuliX for computer vision and OCR, it bridges browser workflows and legacy desktop applications in one script.
Why TagUI instead of UiPath
UiPath approaches robotic process automation through graphical suites like UiPath Studio, proprietary server orchestrators, and Windows runtime environments. While this architecture suits enterprise teams using visual canvases, it introduces friction for developer workflows. UiPath automations rely on complex project packages and XML files that resist clean Git diffs, command-line execution in build pipelines, or unattended runs on headless Linux hosts. Teams automating routine web forms, data scraping, or desktop interactions often find UiPath licensing fees and virtual machine requirements disproportionate to the job.
TagUI fits terminal workflows by storing automation routines in plain text files tracked in Git. Scripts use simple verbs such as click, type, and table that developers can edit anywhere, trigger via shell scripts, or invoke through an official Python package. On continuous integration agents or cron servers, TagUI executes headlessly without desktop login sessions or per-bot fees. In exchange, teams forgo the centralized credential vaulting, visual debuggers, and compliance auditing built into UiPath Orchestrator, handling scheduling and secrets through standard DevOps tooling.
What it does well
- Plain-text scripts in Git: Automation flows live in raw text files, enabling code reviews through standard pull requests instead of proprietary files.
- Headless execution in CI/CD: Scripts run inside headless Chrome on Linux hosts, automating web workflows without a display server.
- Dual selector and visual targeting: Teams can target web elements via XPath and CSS through Chrome DevTools Protocol, or use OpenCV matching and OCR for desktop windows.
- Python and shell integration: The CLI accepts runtime parameters, runs from shell scripts, and offers a Python package to orchestrate steps in data pipelines.
- Direct table extraction: Built-in commands scrape HTML tables into CSV and Excel files without custom parsing glue code.