Mark Castillo
Trust3 Labs is the research arm of Trust3 AI. Our team studies how autonomous AI agents behave in real enterprise environments, where access control, data governance and runtime enforcement break down, and what it takes to secure them. We publish engineering deep dives, threat research, open tools and conference talks so security and data leaders can understand emerging risks such as agent-to-agent communication, MCP access, credential exposure and policy drift, and put practical controls in place before those risks reach production. This page lists the articles written by this author, with the most recent first. Every article is written by the practitioners who build and test the Trust3 AI platform.
Founding AI Solutions Architect
Securing A2A’s Two Attack Surfaces: Ingress vs. Egress
Key Takeaways A2A calls are one more entry point for prompt injection: a receiving agent has no way to tell trusted instruction…
Mark Castillo · October 7, 2026
The 3 Layers of Agent Enforcement: Where Your SSH Key Actually Leaks
TL;DR Scope is pro-code agents (Claude Code, Codex, Cursor, Cline) on a developer’s machine, not no-code platforms. Infrastructure controls don’t apply to…
Mark Castillo · August 17, 2026