Blog
Securing agentic AI and the Model Context Protocol.
stdio tunnels: Enabling Cloud Agents to reach local MCPs
An entire category of MCP connectors is reserved for local CLI agents. stdio tunnels unlock local MCPs, like Beeper for WhatsApp/LinkedIn/iMessage, for any cloud or mobile AI agent.
Why agentic AI needs a data firewall, not another dashboard
AI agents now read and write real company data through MCP. Governing that access at runtime, deterministically, is the gap Sealgate closes.
MCP Security for Security Leaders: Why Local STDIO Servers Are the New Shadow IT
Local STDIO MCP servers are unmanaged code with privileged API access that no DLP, CASB, or EDR inspects. The 2026 CVE record, and how to govern it.
MCP Gateway Comparison for Security Leaders
MCP gateways standardize enterprise AI configuration but leave real observability, data-exfiltration, and multi-tenancy gaps. What security leaders should evaluate before choosing one.
Agentic AI Disrupts Traditional Data Security Posture
Autonomous AI agents inherit their user's access, act across tools, and slip past network monitoring. Why RBAC breaks for agents, and what an agentic gateway fixes.
Subscribe
New writing on securing agentic AI and MCP, straight to your inbox. No spam, unsubscribe anytime.