The internet just got a new kind of visitor.
For thirty years the web assumed a person behind every request. AI agents broke that assumption — and the old tools (CAPTCHAs, IP blocks, robots.txt) were never built to tell one agent from another.
AIdenID is the answer: give every AI agent a verifiable identity, and give every website a clear, provable decision about what that agent may do — on every request. Not "human or bot," but which agent, acting for whom, under what authority — with a signed receipt anyone can check.
SentinelLayer — the stack AIdenID is built on.
AIdenID wasn't built the normal way. It was built on SentinelLayer: a governed operating system for autonomous AI agents that its founder built first, and then used to build the product. Three parts:
Omar Gate
A spec-driven AI security gate on every GitHub pull request: scan → block → remediate → re-review. Nothing merges until it passes.
Senti
The coordination layer where a human and several governed AI agents work the same task together, in shared rooms, with a full audit trail.
The CLI
The command line that drives the whole loop — provision, run, review, and ship — so one operator moves like a team.
One founder. Four shipped artifacts. A team built with intent.
We don't measure ourselves in headcount. We measure ourselves in what's shipped, gated, and provable.
Built AIdenID, SentinelLayer, Senti, and Omar Gate over nine months — solo on the keyboard, paired with governed AI agents through Senti. The four artifacts that frame the founder:
Owns verifier-SDK architecture and design-partner conversion, paired through Senti. Long shared history with Carther on the infrastructure these products are made of.
Shaped AIdenID's early governance and policy architecture; met Carther through MIT Sandbox. Advises on governance and policy design.