---
title: "Three doors to the same conversation: the vault on screen, the AI client you already use, and your team's API."
seo_title: "Interfaces: vault, agent channel and API | SazAI"
description: "The agent answers in your vault, your AI client and your team's API: the same engine, the same vault, open standards."
tags: [marketing]
---

> You do not switch products when you switch places. The question you ask on the vault screen is the same one your AI client asks through the agent channel, and the same one your system asks through the API. What changes is the door; what answers is always the same engine reading the same vault.

<div class="chap">

<div class="ico-col"><sz-icon style="color:var(--verb-talk)" name="chat" size="36"></sz-icon></div>

<h2 id="three-doors">Three doors, one conversation</h2>

<h3 class="sub">On the vault screen, in the AI client you already use, or straight from your system</h3>

On screen, you talk to the agent next to the documents it is reading: you ask, you get the
answer with the source, you click the source and land on the slide. In your AI client, the
vault becomes one more thing that client knows how to query, without leaving where you already
work. And through the programming channel, your team automates from outside exactly what a
person does by talking.

<figure class="viz duo" role="group" aria-label="Three doors around a single vault: the screen, the AI client and the API">
<div class="side"><svg viewBox="0 0 48 48" aria-hidden="true"><rect x="6" y="10" width="36" height="26" rx="3" fill="none" stroke="var(--verb-talk)" stroke-width="1.5"/><line x1="18" y1="40" x2="30" y2="40" stroke="var(--verb-talk)" stroke-width="1.5"/><line x1="24" y1="36" x2="24" y2="40" stroke="var(--verb-talk)" stroke-width="1.5"/></svg><b>the screen</b><span>the agent next to the documents</span></div>
<div class="mid"><svg viewBox="0 0 120 72" aria-hidden="true"><rect x="44" y="22" width="32" height="28" rx="5" fill="var(--verb-talk)" opacity=".14" stroke="var(--verb-talk)" stroke-width="1.5"/><line x1="10" y1="18" x2="44" y2="30" stroke="var(--gray)" stroke-width="1.25"/><line x1="10" y1="54" x2="44" y2="42" stroke="var(--gray)" stroke-width="1.25"/><line x1="76" y1="30" x2="110" y2="18" stroke="var(--gray)" stroke-width="1.25"/><line x1="76" y1="42" x2="110" y2="54" stroke="var(--gray)" stroke-width="1.25"/></svg><b>one vault</b><span>the same engine, the same reading</span></div>
<div class="side"><svg viewBox="0 0 48 48" aria-hidden="true"><path d="M24 8l3.5 10.5L38 22l-10.5 3.5L24 36l-3.5-10.5L10 22l10.5-3.5z" fill="none" stroke="var(--verb-talk)" stroke-width="1.5"/><line x1="12" y1="42" x2="36" y2="42" stroke="var(--lightgray)" stroke-width="1.5"/></svg><b>your client, your system</b><span>through the agent channel or the API</span></div>
<figcaption>Three doors, and behind all three the same vault with the same reading.</figcaption>
</figure>

In all three, the answer arrives the same way: with the address of where it came from, and
with a refusal when the subject is not in your documents.

</div>

<div class="chap">

<div class="ico-col"><sz-icon style="color:var(--verb-talk)" name="globe" size="36"></sz-icon></div>

<h2 id="open-standard">The agent channel is an open standard</h2>

<h3 class="sub">Authentication and discovery by public standards, not by a convention of our own</h3>

The channel through which an AI client talks to your vault is not a house invention. It is a
<dfn tabindex="0" data-hint="The server that holds the resource and requires a credential to deliver it, in the web's standard authorization design. Whoever arrives without a credential gets 401; whoever wants to know how to authenticate reads the discovery document.">resource server</dfn> in the web's standard authorization design: whoever arrives without a
credential gets 401, and the document that says how to authenticate is published for any
client to read on its own. A convention of our own was dropped on purpose, because it would
require every integration to learn our way before asking the first question.

<figure class="viz" role="group" aria-label="The agent endpoint answers 401 without a credential and publishes its discovery">
<div class="ev live"><span class="n">401</span><span class="t"><b>The agent channel is live</b>The MCP endpoint answers 401 without a credential, which is the correct behavior for an OAuth resource server, and discovery at .well-known answers 200. That is the machine talking, not us.</span></div>
<figcaption>A 401 here is proof the door exists and is locked, not that it fell.</figcaption>
</figure>

</div>

<div class="chap">

<div class="ico-col"><sz-icon style="color:var(--verb-talk)" name="share" size="36"></sz-icon></div>

<h2 id="one-substrate">One substrate, several readers</h2>

<h3 class="sub">A new channel does not invent a new surface</h3>

An open API and an agent channel are not plumbing: they are what lets other interfaces get
built on the same substrate, without us in the room. A third-party agent, a panel for your
team, a workflow integration. The product gets to be those things because the doors exist,
not because we foresaw each one.

And what keeps the doors from turning into different products is a rule that holds as a
contract: the set of operations exposed to the agent is a subset of what the API exposes.
Nothing exists in one channel without existing in the other, so a new capability does not
appear on one side and vanish on the other. The whole set, and what it refuses, is in
[capabilities](capabilities).

</div>

<div class="chap">

<div class="ico-col"><sz-icon style="color:var(--verb-talk)" name="folder" size="36"></sz-icon></div>

<h2 id="same-vault">The door changes; the owner of the vault does not</h2>

<h3 class="sub">The conversation stays with you, wherever you speak from</h3>

From any of the three doors, the vault is yours: the same isolation between organizations, the
same metering, the same record of who did what. The conversation stays with you, and nothing
the agent answers is built with another company's data, because it has no source beyond your
vault.

The boundary of this sentence gets stated here and not in a footnote: the text that goes to
the model follows the provider's retention terms, and the agent's own first line already says
so. What we do with your data, and what you can audit on your own, is in
[privacy](privacy).

</div>

These are the doors. What passes through them, operation by operation, is in
[capabilities](capabilities), and the reason the file is never touched by the model is in
[the conversation is with the AI](docs-by-spec).
