Cursor Launches Automations: Always-On AI Agents for Code Review, Security, and Incident Response

Cursor, the AI-powered code editor with over $2 billion in annual revenue, launched Automations on March 5, 2026 — a system that automatically triggers coding agents based on events like code commits, Slack messages, PagerDuty incidents, or scheduled timers.

The move marks a significant shift from the "prompt-and-monitor" model that has defined agentic coding so far.

What Are Cursor Automations?

Automations let engineers configure AI agents that run without human initiation. Instead of manually prompting an agent for each task, developers set up triggers:

Code changes — an agent reviews every push to main for bugs, security issues, or style violations
Slack messages — a message in a specific channel kicks off an automated workflow
Timers — weekly codebase summaries, dependency checks, or changelog generation
PagerDuty incidents — an agent immediately queries server logs via MCP and proposes a fix
Custom webhooks — any external event can trigger an automation

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

Three Categories of Automations

1. Review and Monitoring

Cursor's existing Bugbot — which catches bugs on every PR — is the prototype. Automations expand this to:

Security audits triggered on every push to main, posting high-risk findings to Slack
Agentic codeowners that classify PR risk by blast radius and complexity, auto-approving low-risk changes and assigning reviewers for high-risk ones
Deeper analysis — spending more tokens to find more nuanced issues

2. Incident Response

When a PagerDuty incident fires, an automation spins up an agent that:

Queries Datadog logs via MCP
Inspects recent codebase changes
Posts findings to Slack with a proposed fix PR

Cursor reports this has "significantly reduced" their own incident response time.

3. Chores and Knowledge Work

Everyday tasks that previously fell through the cracks:

Weekly codebase change summaries posted to Slack
Dependency updates and changelog generation
Documentation sync after major changes

Why This Matters

The agentic coding landscape is intensifying. OpenAI updated Codex with dedicated hardware, and Anthropic added voice mode to Claude Code. Cursor's answer is to make agents always-on rather than always-prompted.

This addresses the core bottleneck: as engineers oversee dozens of agents simultaneously, human attention becomes the limiting resource. Automations turn engineers from agent operators into system designers who configure workflows and intervene only when needed.

Cursor runs hundreds of automations per hour internally. Each automation spins up a cloud sandbox, follows configured instructions with chosen models and MCPs, and verifies its own output — with a memory tool that lets agents learn from past runs.

🚀 Building with AI coding agents? Noqta helps teams implement agentic development workflows that actually ship — from setup to production.

How It Compares

This launch builds on Cursor's cloud agents (February 2026), which run autonomously for up to 52 hours. Automations add the trigger layer — the ability to launch those agents without human prompting.

Compared to Claude Code's CI/CD integration or GitHub's Agent HQ, Cursor's approach is more event-driven and IDE-native. The competition between coding tools continues to accelerate.

Market Context

Revenue: Cursor's annual revenue doubled to over $2 billion in three months (Bloomberg, March 2)
Market share: ~25% of generative AI clients subscribe to Cursor (Ramp data)
Competition: OpenAI Codex, Anthropic Claude Code, GitHub Copilot, and Windsurf all competing for the agentic coding market

💡 Need your AI-generated code reviewed before it hits production? Noqta's audit service catches what automated agents miss — security gaps, architecture drift, and technical debt.

What's Next

Cursor is positioning Automations as the foundation for what they call the "software factory" — where AI agents continuously monitor, review, and improve codebases with humans overseeing the system rather than directing every action.

For engineering teams evaluating their AI coding stack, Automations raises the bar on what "agentic" means: not just AI that writes code on command, but AI that maintains, reviews, and responds to your codebase 24/7.

FAQ

Is Cursor Automations available to all users?

Automations is rolling out now. Users can configure automations at cursor.com/automations or start from templates.

How much does Cursor Automations cost?

Pricing details haven't been announced separately — it's part of the Cursor platform. Check our pricing comparison for how Cursor stacks up against alternatives.

Can Automations work with external tools?

Yes. Automations support MCP connections to tools like Datadog, PagerDuty, Slack, GitHub, Linear, and Notion. Custom webhooks allow integration with any external service.

How is this different from CI/CD pipelines?

CI/CD pipelines run predefined scripts. Automations launch AI agents that can reason about code, make decisions, and produce fixes — not just run tests.