Coding Assistants: Minimizing Risk Where Agents Take Action

Coding assistants have gone from experiment to critical infrastructure faster than most security teams expected. While tools like Claude Code, Cursor, and GitHub Copilot drive real productivity for development teams, the permissions they inherit and the autonomous decisions they make open up risks that sit almost entirely outside the defenses security teams have built in the last decade. Join Zenity Labs' research team live as they unpack the full coding agent threat model and the gaps most teams aren't watching for yet.

Register for our July 9 webinar to learn:

• Why coding assistants have become the predominant enterprise agent use case
• Why coding assistants’ abilities to read/edit code, invoke tools, and access MCP servers create a persistent and exploitable threat vector
• Where traditional EDR/XDR controls fall short for endpoint agents, and how the combination of high privilege and low visibility leaves the door open for malicious actors
• How to configure, continuously monitor, and resolve incidents in coding assistants living on the endpoint.

Greg Zemlin

Technical Product Marketing

Tamir Ishay Sharbat

Keren Katz

Director of Threat Detection Research