Cursor Enterprise

Agent skill enabling iterative looping and refinement of code generation tasks.

Core agentic coding interface that plans, builds, debugs, and reviews code across desktop, CLI, web, and mobile surfaces. Supports multi-model subagents running in parallel.

Feature enabling agents to automatically review code changes before implementation, providing validation and quality checks.

Always-on agents that run on schedules or triggers to automatically build, maintain, and fix software including CI failure remediation.

Remote compute infrastructure allowing agents to build, test, and demo features autonomously without consuming local resources.

Custom embedding model that enables agents to achieve best-in-class recall across large codebases with semantic search capabilities.

A code composition tool with command (/) and file (@) shortcuts that assists with multi-step code generation and editing tasks within Cursor.

System using @-mentions and image uploads to point agents at exactly what matters for reference and context injection.

An AI coding agent that autonomously plans, searches, builds, and deploys code features end-to-end. Runs on its own compute to build, test, and demo software for human review.

Autonomous AI agents that build, test, and demo features end-to-end, running in parallel on their own compute with the ability to work across codebases.

Command-line interface for Cursor Agent, enabling developers to invoke AI-powered code assistance and planning from the terminal.

Cloud-based agents that autonomously execute tasks, build, test, and deploy features without requiring local compute resources.

Code review automation capability that provides intelligent feedback and PR analysis.

AI-powered code editor for manual to agentic coding workflows with integrated agent capabilities and planning features.

Integration that brings Cursor agent capabilities into Slack, enabling teams to request code changes and collaborate on software development directly within their chat workspace.

Integration connecting Cursor Agent to Jira for issue tracking and automated development workflows.

Integration enabling teams to invoke Cursor Agent for code generation, planning, and PR review directly within Slack channels.

Mobile application enabling developers to access Cursor's AI coding capabilities from iOS devices.

A specialized autocomplete model that predicts the next code action with high accuracy and speed. Offers fine-grained control over AI autonomy levels.

Cloud-based agent execution accessible from browser or mobile device with cloud-hosted agent dashboard and task management.

Agent mode that instruments code and uses real execution data to pinpoint and fix bugs with targeted solutions.

Visual prompt interface allowing developers to direct agents using design-based inputs and visual specifications.

Code versioning and snapshot rollback system showing evolution of code with ability to revert to previous states anytime.

Protocol to connect external tools and data sources like GitHub and Figma directly to Cursor agents.

Model Context Protocol integrations and organizational marketplace for sharing tools and capabilities across teams.

Grid view of all open windows as scaled previews with spring-based animations, allowing quick window selection and management.

Agent mode for complex tasks that asks clarifying questions, builds a plan, and executes in the background.

Community-built plugin system to extend Cursor with new capabilities from external tools and integrations.

Collaborative feature allowing multiple users to work together on code in shared editing spaces.

Feature allowing parallel conversations within Cursor and search across conversation history.

Domain knowledge layer allowing Cursor to discover and run specialized prompts and code for team-specific tasks.

Agent capability to review pull requests and perform coding tasks directly from Slack messages.

Multiple AI agents that run in parallel to explore codebases, with each using the best model for its specific task.

Organizational marketplace functionality enabling teams to share and manage Model Context Protocols (MCPs) and tools across Cursor deployments.

System to teach Cursor team preferences, conventions, and specific architectural decisions for consistent code generation.

Direct shell command execution from Cursor for builds, tests, installs, and other terminal operations with sandboxed defaults.