About
An agent cannot approve itself.
The model proposes. The gate authorises.
Sector8 places an independent decision gate between agents and real-world actions. Every tool call, API call, model call, and handoff can be allowed, denied, or audited before execution.
Founders
Kammil Mahajan and Liban Mohamud are co-founders and cybersecurity practitioners. Both studied at University College Dublin between 2011 and 2013, where they earned an MSc in Digital Investigation and Forensic Computing.
After years in enterprise security, compliance, and large-scale systems, they reconnected around a concrete gap: organisations were putting agents into production that could approve their own actions.
Sector8 is the independent gate. Shadow mode first. Fail-closed at the gate.
The foundation
Our mission
The name is the company. The product is the gate. We do not try to govern what the model thinks. We govern what it is allowed to do.
1
Propose
The agent emits a tool call, API call, shell command, or handoff. Before execution, Sector8 receives the proposed action.
2
Decide
A deterministic policy engine returns ALLOW or DENY. If it cannot allow the action safely, it denies. The agent does not approve itself.
3
Prove
Every decision carries a reason code, decision trace ID, evidence hash, and policy version. Allows are recorded, not just denials.
How we work
What we stand for
Actions, not answers
A wrong sentence is embarrassing. A wrong tool call moves money, deletes records, or exposes data.
Determinism
Same action, same policy, same decision. If it is not reproducible, it is not a control.
Evidence
A denial without a name is a black box. Every decision leaves a record.
Shadow mode first
Enforcement off. A report of what the agents already tried. Then you choose what to enforce.
Fail-closed
If the gate cannot decide, the action does not run — for actions your app sends to Sector8.