Not every action deserves the same approval. A read-only lookup risks nothing. A routine update risks a little. A consequential action risks a lot.
Most approval flows ignore that. They follow the org chart instead of the risk. Everything waits for the same manager. Everything takes the same two days. The low-risk work stalls, and the high-risk work gets the same rubber stamp as everything else.
Practitioners wrote the real rules for us, and none of them are in any API contract. Testimonials need a human before they go live. Escalation happens after two contacts. Refunds under fifty dollars are fine when the order is paid and never refunded before. Timeline rules nobody wrote down anywhere.
Those are not edge cases. Those are the approval flow. The risk lives in the unwritten rule, not in the endpoint. An agent that cannot see the rule cannot follow it, no matter how clean the permission model looks.
Design approvals around the risk. Let the read go through. Let the routine update go through with a log. Make the consequential action wait for a named human, with the reason attached.
What is the highest-risk action your agent can take today without asking anyone.
Top comments (0)