AI-Powered RPA
Plain English in, desktop automation out.
Describe a task in plain English and this engine turns it into a validated, replayable workflow that clicks, types and navigates your desktop. Dangerous actions are blocked before they run.
- Tests passing
- 186
→Natural language to strict JSON workflows
→Safety engine blocks destructive commands by default
→Dry-run, record-and-replay, wait-for-element
Two ways to read it
In plain English
Describe a task in plain English and this engine turns it into a validated, replayable workflow that clicks, types and navigates your desktop. Dangerous actions are blocked before they run.
For engineers
Python + Pydantic v2 workflows driving pyautogui/pynput, with a safety engine, dry-run mode, record-and-replay and redacted audit logs.
What it does.
- Natural language to strict JSON workflows
- Safety engine blocks destructive commands by default
- Dry-run, record-and-replay, wait-for-element
Built with
- Python
- Pydantic v2
- pyautogui
- pynput
Status: Shipped · started January 2026 · open on GitHub
Keep exploring
- AI agents & orchestrationShipped
Iron Jarvis
Your own local-first AI operating system.
59★GitHub stars
- AI agents & orchestrationExperiment
RealDeal CPA Agent
An AI accounting employee, not a chatbot.
117Skills
- AI agents & orchestrationShipped
BetaBot
An AI beta tester that files real bug reports.
70Tests passing