AIR has emerged from stealth with a security platform focused on the software supply chain around AI agents: skills, plugins, MCP servers, sub-agents and the external context those systems consume.
The company says its platform discovers agents across an organization, evaluates the components they depend on, filters untrusted context before it reaches an agent and monitors actions at runtime. It also maintains a marketplace of vetted add-ons.
That matters because an agent component can change after installation. A skill or MCP server that looked safe when approved may later fetch different instructions, change ownership or depend on a compromised external resource. AIR is therefore selling continuous verification rather than a one-time scan.
The company has raised $50 million across two seed rounds led by Sequoia and Greenoaks. Its claims about the effectiveness of its controls still need independent testing, but the launch reflects a real shift: agent security is becoming a separate enterprise control layer.