Snyk acquired Invariant Labs, an ETH Zurich spin-off focused on AI-agent security, to accelerate work on MCP vulnerabilities, tool poisoning and Guardrails, the company said on 24 June 2025.
Developer security vendors are extending scanning from code and containers into agent tools and MCP-style connectors. AppSec buyers still need reproducible findings on their own agent stacks.
Snyk already sells a developer-first platform across code, open source, containers and IaC. Invariant Labs is a bid to keep agentic security research inside that product line as AI-assisted software development spreads.
Security leaders should demand Guardrails evidence on their MCP and tool inventory — before treating any agentic AI security acquisition as a completed AppSec programme.