Cursor Launches 'Always-On' AI Coding Agents
The AI-native code editor Cursor has launched a new 'Automations' feature, enabling always-on AI agents that trigger on events like GitHub PRs or Slack messages. The move points toward a future where tasks like code review and CI/CD are increasingly handled by autonomous agents.
Cursor's "Automations" feature moves beyond simple code suggestions by allowing AI agents to operate independently in the background. These agents are triggered by events such as a new GitHub pull request, a Slack message, or on a recurring schedule, enabling them to perform tasks without manual prompting. This shifts the paradigm from an "agent in your editor" to an "agent in your pipeline." Use cases for these always-on agents include automatically reviewing code for vulnerabilities, triaging bug reports, responding to PagerDuty incidents by investigating logs and proposing fixes, and even generating weekly Slack summaries of repository changes. The agents run in isolated cloud sandboxes and can be configured to use specific AI models and tools, learning from past runs to improve their accuracy over time. The company behind Cursor, Anysphere, was founded in 2022 by four MIT graduates: Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger. Their AI-native code editor, a fork of VS Code, quickly gained traction in the developer community. Cursor has seen explosive growth, reaching a $29.3 billion valuation after a $2.3 billion funding round in November 2025. The company has surpassed $1 billion in annualized revenue and is backed by major investors including Accel, Coatue, Nvidia, and Google. This move toward autonomous agents reflects a broader industry trend of integrating AI more deeply into the software development lifecycle (SDLC). Unlike passive AI assistants like early versions of GitHub Copilot that wait for user prompts, these agents are designed to take initiative, orchestrate complex workflows, and manage multi-step processes with minimal human intervention. The goal is to automate repetitive work, allowing developers to focus on more complex architectural and strategic tasks.
