About

I am a security researcher focused on hardware attestation and confidential computing, especially the boundaries where software makes a security decision. I also study how AI agents and MCP servers receive and use authority.

I read specifications and source code together, then trace each security claim to the code that parses evidence, applies policy, and authorizes an action. A verifier can accept valid evidence, but the system can still fail if it checks the wrong property or skips a path. I reproduce each issue locally, try to disprove it, and report only the cases that survive.

My public work includes two Intel TDX Guest CVEs and a vulnerability in the Microsoft Open Enclave SDK. My reports and patches have also led to fixes in the Linux kernel, DPDK, Fortanix software, Tokio, and tokio-vsock.

I also build security research tools, attestation libraries, and technical specifications. I write and maintain AIR, an active individual IETF Internet-Draft for attestation-linked AI inference receipts. My public code is on GitHub.

If you work on attestation, confidential computing, or agent security, send me an email.