Cursor announces major update to AI agents as coding tool battle heats up
Summary
Cursor updated its AI coding agents to test their own changes and record work through videos, logs and screenshots. The agents can be triggered from the web, Cursor's desktop app, mobile devices, Slack or GitHub and can run on parallel cloud-based virtual machines to avoid local resource contention and speed onboarding. The update enables agents to run many tasks in parallel, automate testing and iterate on features so developers spend less time editing code and more time on judgment and design. Cursor reports roughly 35% of its pull requests are now generated by agents running on their own VMs, and the move positions the company to better compete with Anthropic, OpenAI and Microsoft in the AI coding tools market.
Classifications
industries
Fintech & Banking
applications
Accounting and Taxes
AskAI Classifications
Labels
SaaS
Cloud Computing
Enterprise Software
