Cursor launches ‘Automations’ to let AI coding agents run tasks independently

The growing presence of artificial intelligence in software development is gradually transforming how programmers work. Instead of writing every line of code themselves, many engineers now collaborate with AI-powered tools capable of generating, reviewing and analysing code. However, as more AI assistants become part of the development environment, developers are also encountering a new challenge — managing and coordinating multiple AI systems simultaneously.

To address this issue, Cursor has introduced a new feature called Automations, designed to allow AI coding agents to initiate tasks automatically within a development workflow. The system aims to reduce the need for engineers to manually prompt and monitor several AI tools throughout the day.

In many AI-assisted coding environments today, developers follow what is commonly described as a “prompt-and-monitor” workflow. Engineers instruct an AI system, examine its output and then provide the next prompt. While this process can speed up development, it also requires constant attention. As AI agents become capable of handling more advanced tasks, managing multiple prompts across different tools can quickly become overwhelming.

Cursor’s Automations feature attempts to streamline that process by allowing AI agents to begin work automatically when certain triggers occur. For example, an automation can start when new code is added to a repository, when a message appears on Slack, or when a scheduled timer activates.

Once triggered, the AI agent can carry out tasks such as reviewing newly added code, analysing project changes, or checking for potential issues. Developers are then alerted only when human input or approval becomes necessary.

“It’s not that humans are completely out of the picture,” Jonas Nelle, Cursor’s engineering chief for asynchronous agents, told TechCrunch. “It’s that they aren’t always initiating. They’re called in at the right points in this conveyor belt.”

The new feature builds on tools already available within Cursor’s ecosystem. One such tool is Bugbot, which automatically scans newly added code for possible errors whenever developers update the codebase. Instead of requesting checks manually, the system works in the background and alerts engineers when problems are detected.

With Automations, Cursor says similar capabilities can now extend to more advanced workflows. AI agents can perform deeper security assessments, analyse broader sections of a project and identify issues without constant supervision.

Engineering lead Josh Ma said allowing AI systems more time and computing power to analyse code can significantly improve their ability to detect complex problems. “This idea of thinking harder, spending more tokens to find harder issues, has been really valuable,” he said.

Cursor has already implemented the automation system internally for several operational tasks. For instance, when a technical incident is reported through PagerDuty, an AI automation can immediately check server logs through an MCP connection. Another automation compiles weekly summaries of codebase updates and posts them on the company’s Slack channels.

According to the company, hundreds of such automated processes are executed every hour.

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

The launch comes amid a rapidly expanding market for AI coding tools. Companies including OpenAI and Anthropic are also introducing systems designed to enable AI agents to manage larger portions of the software development process.

Despite rising competition, Cursor appears to be growing quickly. Data from Ramp suggests that about 25 percent of generative AI clients currently use Cursor in some capacity. Meanwhile, a recent report by Bloomberg indicated that the company’s annual revenue has surpassed $2 billion, doubling within the past three months as demand for AI-powered coding tools continues to surge.