Cursor's New AI Agent Transforms Coding

The proliferation of agentic coding has significantly escalated the complexity of a software engineer's daily work. Modern engineers may find themselves managing dozens of coding agents concurrently, initiating and directing various processes as required. This intricate oversight demands considerable attention, quickly rendering human engineers' focus the primary limiting factor in productivity.

To address this burgeoning complexity, Cursor is introducing a new solution today called Automations. This innovative system empowers users to automatically launch agents within their coding environment. These launches can be triggered by events such as new additions to the codebase, incoming Slack messages, or predefined timers. Cursor positions Automations as a sophisticated method to review and maintain the extensive code generated by agentic tools, eliminating the need for engineers to manually track numerous agents simultaneously.

Fundamentally, Automations enable engineers to transcend the conventional "prompt-and-monitor" paradigm that characterizes much of agent-based engineering. Rather than requiring a human to manually prompt agents for every task, Cursor's Automation framework facilitates autonomous agent activation, ensuring human involvement is strategically integrated only when essential.

"It's not that humans are completely out of the picture," Jonas Nelle, one of Cursor's engineering chiefs for asynchronous agents, clarified to TechCrunch. "It's that they aren't always initiating. They're called in at the right points in this conveyor belt."

A precursor to this broader Automation system is Bugbot, a long-standing Cursor feature. Bugbot activates each time an engineer commits new code to the codebase, meticulously reviewing it for bugs and other potential issues. By leveraging Automations, Cursor has successfully enhanced this system, extending its capabilities to encompass more thorough security audits and comprehensive review processes.

Engineering lead Josh Ma remarked, "This idea of thinking harder, spending more tokens to find harder issues, has been really valuable."

Cursor estimates that its platform currently executes hundreds of automations per hour, extending far beyond routine code review. The system also plays a crucial role in incident response, where PagerDuty incidents can trigger an agent to instantly query server logs via an MCP connection. Furthermore, a separate automation provides weekly summaries of codebase changes directly to Cursor's company Slack channel.

"In the abstract, anything that an automation kicks off, a human could have also kicked off," Nelle explained. "But by making it automatic, you change the types of tasks that models can usefully do in a code base."

This launch occurs amidst an intensely competitive landscape within the agentic coding sector. Both OpenAI and Anthropic have unveiled significant updates to their respective agentic coding tools in the past month, intensifying the race for innovation.

Despite the competitive environment, data from Ramp indicates that Cursor's market share has remained stable since May, with approximately 25% of generative AI clients subscribing to Cursor in some capacity.

Nonetheless, the overall expansion of the agentic coding space has fueled a remarkable surge in Cursor's revenue. Bloomberg reported earlier this week that Cursor's annual revenue has soared to over $2 billion, effectively doubling within the last three months alone.