Varonis Atlas adds Cursor support for runtime enforcement and session forensics

Varonis Systems, Inc. said Varonis Atlas added support for Cursor, an AI-native coding tool, to provide controls for agent-driven work inside codebases. The update addresses access and enforcement needs when coding agents read, write, and execute through development workflows that can touch sensitive material.

Varonis Atlas is positioned to apply security controls across the agentic development lifecycle for Cursor, using runtime enforcement, threat detection, and full session forensics. Varonis also tied the Cursor support to broader coverage of agentic IDE environments and related AI security capabilities.

According to the release, Cursor agents read, write, and execute inside a codebase by running terminal commands, installing dependencies, and calling MCP-connected tools. The company said this creates access to source code, .env files, credentials, API keys, and customer data, and that Atlas focuses on securing sensitive data via runtime enforcement and threat detection. Atlas was described as supporting runtime enforcement, session forensics and threat detection, sensitive data and secrets protection, and shadow AI discovery.

Ron Bennatan, VP of AI and Data Security Strategy at Varonis, said, “The agents may change, but the security requirements don't,” adding that Atlas provides a single place to secure data access, monitor agent activity for risky actions, and enforce policy. The release also described an earlier integration with the Claude Compliance API for Claude Enterprise and Claude Platform and said the Cursor work continued coverage for agentic IDE tools such as Claude Code, GitHub Copilot, and VS Code. The company said the support provides one set of controls regardless of which tools and vendors agentic AI work uses.

Provided by Globe Newswire on behalf of Varonis. Click to read original content.