Cursor launched Automations on Thursday, a framework that lets coding agents run continuously without anyone typing a prompt. Triggers include scheduled timers, Slack messages, new GitHub PRs, Linear issues, and PagerDuty incidents. Custom webhooks cover anything else.
When triggered, an agent spins up a cloud sandbox, uses whatever MCPs and models you've configured, and verifies its own output. A memory tool lets agents learn from past runs. Cursor says it already runs hundreds of automations per hour on its own codebase, including a security review that fires on every push to main and a weekly Slack digest of codebase changes. Jonas Nelle, Cursor's engineering chief for asynchronous agents, told TechCrunch: "They're called in at the right points in this conveyor belt" -- meaning humans don't disappear, they just stop initiating every task. That's a reasonable reframe, though how reliably agents self-verify at scale is still an open question.
The feature has a clear ancestor: Bugbot, Cursor's existing PR review tool, operates on the same trigger-and-run logic and reportedly fires thousands of times daily. Automations generalize that pattern to incident response, risk triage, and routine maintenance. Start at cursor.com/automations or pick a template.
The launch arrives as Bloomberg recently reported Cursor's annualized revenue surpassed $2 billion, doubling over roughly the past quarter. Ramp data puts Cursor at about 25% market share among generative AI software buyers -- holding steady despite pressure from OpenAI and Anthropic both shipping agentic coding updates in recent weeks.
Bottom Line
Cursor Automations went live March 5 with support for Slack, GitHub, Linear, PagerDuty, and custom webhooks; Cursor reports running hundreds of automations per hour internally.
Quick Facts
- Launched: March 5, 2026
- Triggers: Slack, GitHub PR, Linear issue, PagerDuty, scheduled timers, webhooks
- Cursor runs hundreds of automations per hour (company-reported)
- Cursor annualized revenue: $2B+, doubled in ~3 months (Bloomberg, unverified by Cursor)
- Market share: ~25% of generative AI software buyers (Ramp data)
