AI Guard helps secure coding agents in developer workflows against prompt injection, backdoor attacks, and other core OWASP Top 10 threats. We are building capabilities to detect and block malicious skills, local/uploaded scripts, configurations, and packages directly inline with the coding agent.
Join the Preview of AI Guard for Coding Agents to learn more from our product team and integrate agentic runtime protection in your developer environments, so you can ship AI fast without compromising security. Support for Claude Code is in Preview, with Codex, Cursor, Github Copilot to follow.