Cybersecurity company Snyk Ltd. today announced the launch of Evo by Snyk, an agentic security orchestration system designed to secure artificial intelligence-native applications and tools, including generative and agentic AI.

Snyk is pitching the new Evo offering as ushering in what it calls the “era of the empowered AI Security Engineer.” The AI can act as both guide and autonomous teammate for deploying intelligent orchestration, automation and proactive governance that is required for the security complexities of the agentic enterprise.

Evo by Snyk integrates with and builds upon the Snyk AI Security Platform, providing enterprises with a unified, agentic experience to discover, test, govern and protect AI systems across the entire software development lifecycle.

“Security can’t just keep pace — it must lead the charge,” said Chief Executive Peter McKay. “With Evo, we’re not just securing code — we’re architecting the trust that will unlock the full potential of tomorrow’s agentic systems for the modern enterprise.”

Evo offers agentic AI security that can orchestrate multiple agents, automate workflows and enforce proactive governance across the AI development lifecycle. Using Evo, teams can discover AI components, analyze contextual risk, prioritize actions and create policies to govern AI adoption at scale.

Key features of the new offering include Intelligent Agent Orchestration and Autonomous Task Agents.

With the Intelligent Agent Orchestration feature, a Workflow Agent acts as an intelligent coordinator that combines multiple specialized Task Agents into automated workflows from a single natural language prompt. The Workflow Agent transcends all agent usage in an enterprise, orchestrating both Snyk and non-Snyk agents alike.

Autonomous Task Agents offer a range of specialized Snyk agents to secure every stage of the AI lifecycle. The agents include a Discovery Agent that automatically maps all AI models, application programming interface and dependencies to provide a complete view of AI usage. There’s also a Threat Modeling Agent that automatically builds live AI threat models from code and flags risks like prompt injection with clear remediation paths.

In addition, a Red Teaming Agent runs autonomous adversarial testing of models, agents and applications via the industry’s leading engine for large language model-native application scanning. And finally, an MCP Scan Agent that provides full visibility into all Model Context Protocol servers in developer environments, leveraging Snyk’s MCP research to monitor usage and enforce real-time guardrails.

Other agents offered include an AI Risk Registry Agent that continuously evaluates and scores AI component risk, analyzing security, compliance and data controls to ensure only trusted models and MCP servers are deployed. A Policy Agent defines and enforces executable AI security guardrails for model use, data access and compliance. And a Fix Agent automatically resolves AI security issues through direct remediation or by initiating pull requests.

Additional features of Evo by Snyk include Natural Language Policy Creation, with a Policy Agent that allows teams to proactively create and enforce security policies using natural language. It governs the use of AI in development and runtime applications and Comprehensive Reporting via a Reporting Agent that generates customizable insights across all agents, enabling faster and more flexible AI security risk reporting.

Evo by Snyk is available in preview today for customers, with broader availability expected in early 2026.

Image: SiliconANGLE/Ideogram