Series
Agent approvals
Choose which agent actions can finish automatically and which need a human decision.
4 articles
Reading list
In this series
- 01
· Approvals & policy · 8 min
Designing approvals operators actually read
The failure mode every agent-approval system hits is rubber-stamp clickwrap. Design moves that delay it: signal selection, payload composition, deadlines, audit, and the hand-off problem.
- 02
· Approvals & policy · 8 min
Policy at action time: what the gate evaluates
The five inputs and three checks a runtime policy gate needs to decide whether an agent action should run, with policy-language and approval-UX considerations.
- 03
· Approvals & policy · 3 min
Controls that hold when the agent stops agreeing
An instruction lives inside the agent's context and competes with everything else there. A gate lives outside it, at a boundary the agent cannot cross without a decision it doesn't control.
- 04
· Approvals & policy · 5 min
The model is not the institution
Intelligence and authority are separate properties. Capability decides what work a model is useful for. The system decides what it is allowed to do and what evidence is required before the result is accepted.