← All Jimmy Explains videos
Permission Is Not Proof · 2:26

I’d Fire an Employee for This: Three Checks Before an AI Agent Acts

An agent with valid permissions sends a quote from an expired price sheet. Jimmy walks through three checks: show the evidence, bind approval to the exact decision, prove the effect.

The short answer

Valid permissions do not make an agent’s decision right. Before an agent acts on anything that matters: show the evidence and confirm it is still valid, bind the approval to the exact decision so any change to amount, attachment or recipient voids it, and prove the effect with a receipt from the receiving system. A timeout is never a reason to send twice.

By Jimmy Standaert · Founder, RiGi Group / WhiteWater.ai
Published November 2, 2026

The practical takeaway
Before an agent acts, make the decision reviewable. After it acts, verify the effect.

Questions this video answers

Why aren’t valid agent permissions enough?

Permissions confirm the agent may act. They say nothing about whether the evidence behind this decision is current. In the example, every permission passes and the price sheet expired yesterday.

What does it mean to bind approval to the decision?

The approval covers a specific package: evidence, recipient, attachment and terms, within a defined scope and time window. If the agent changes any of it after approval, the action is checked again before execution.

What is a Governed Decision Object?

The reviewable record that keeps the decision, its supporting evidence, its authority and the exact intended action together.

How do you prove an agent’s action worked?

A success message from the tool is not proof. Record the receiving system’s own receipt and read back the state. If completion is uncertain, reconcile before retrying.

Full video transcript

Read the transcript

Your agent has permission. Its pricing is out of date. Here’s the check I would make before it sends that quote. Imagine a customer asks for updated pricing. The agent can access the CRM, draft the quote, and send the email. All three permissions are valid. But the price sheet it used expired yesterday. That’s the problem behind Permission Is Not Proof. First, make the evidence visible. Which price sheet supports this quote? Who issued it? When was it valid? If it’s stale, missing, or contradicted, pause the consequential action and resolve the gap. Second, bind approval to the specific decision. The person reviewing the quote needs the evidence, uncertainty, recipient, attachment, and proposed terms. An approval should cover that package within a defined scope and time window. Now watch what changes. If the agent edits the amount, swaps the attachment, or changes the recipient after approval, the earlier approval may no longer apply. Our proposed contract requires the action to be checked again before execution. We call the reviewable record a Governed Decision Object. Think of it as the decision, its supporting evidence, its authority, and the exact intended action, kept together. Third, prove the effect. A successful tool response is useful evidence, but the receiving system needs to confirm what actually happened. Record the source-native receipt and read back the relevant state. If completion is uncertain, reconcile before retrying. Don’t turn a timeout into a duplicate send. The paper’s appendix makes these boundaries concrete with an illustrative decision object and six failure cases. Expired evidence, changed recipients, revoked authority, and missing receipts each have a defined expected response. Those are proposed acceptance criteria. They are not executed tests or production proof. This is the direction behind Whitewater AIOS: connecting decisions, evidence, authority, action, and receipts across the AI tools you use. LLM-agnostic. Model-agnostic. Agent-agnostic. The human belongs inside that decision process, with enough context to challenge, approve, narrow, or stop the action. Human decision integrity has to survive delegation. Here’s the practical rule: before an agent acts, make the decision reviewable. After it acts, verify the effect. Read Permission Is Not Proof in the PDF attached to my LinkedIn post. The compact technical appendix is included. Keep the decision visible. Bring back the receipt.

Download transcript · English captions

Sources and related reading

More Jimmy Explains

Bring one workflow to the conversation. Jimmy@RiGiGroup.com
rigigroup.com · WhiteWater.ai