Cursor launches iPhone app for overseeing coding agents

Cursor released an iPhone app on June 29 that lets users prompt coding agents from a phone, spin up new agents, and check on agents that were started on desktop.

It is free with in-app purchases and built only for iPhone. It lets users kick off as many agents as they want in parallel, track long-running work, review changes, ship pull requests, inspect diffs, and test code in sandboxed environments. It also supports generated artifacts such as videos, screenshots and logs, plus image annotations with circles and comments for visual feedback and a voice mode for conversation.

Cursor announced its web and mobile agent product on June 30, 2025. That earlier system already let developers run tasks in the background, reach agents from desktop, tablet or mobile browser, review diffs and pull requests, attach images, run multiple agents at once, and receive Slack notifications or trigger work through @Cursor mentions.

Cursor introduced Origin, a Git platform designed for both humans and AI agents, and Michael Truell said a first fully self-trained in-house model was already underway and expected to ship in the next few weeks. Origin was already being used internally and with select partners, with broader availability planned for fall.

Cursor’s App Store listing also said its systems can use frontier models from OpenAI, Anthropic, Google and others, while its own agentic model, Composer, powers the product. Cursor’s move landed in a competitive moment as Anthropic and OpenAI have also moved into mobile tooling for agent oversight.