The control layer for enterprise AI agents

Gatehouse is the control layer Surehand built for AI agents that act on enterprise systems and data. Your team writes the rules once in a signed file: which systems the agent may touch, what it may spend, what needs a person. Every action is checked against those rules before it runs, and anything not allowed is refused. Risky calls and anything over a spend limit wait for one named approver, with the evidence gathered. Every action, hold and decision goes into a record chained with SHA-256, so any edit shows, and it exports in open formats your auditors can read without Gatehouse. Runs in your cloud or Surehand's. A simulated payment run, with the hash chain computed in your browser, is at gatehouse.run.